CAN w PUBLIC DIPLOMACY SURVIVE THE INTERNET? BOTS, ECHO CHAMBERS, AND DISINFORMATION Edited by Shawn Powers and Markos Kounalakis May 201 7 TRANSMITTAL LETTER To the President, Congress, Secretary of State and the American People: Established in 1948, the U.S. Advisory Commission on Public Diplomacy (ACPD) is authorized pur¬ suant to Public Law 1 14- 1 13 to appraise all U.S. government efforts to understand, inform and in¬ fluence foreign publics. We achieve this goal in a variety of ways, including, among other efforts, offering policy recommendations, and through our Comprehensive Annual Report, which tracks how the roughly $1 .8 billion in appropriated funds is spent on public diplomacy efforts throughout the world. Part of the Commission’s mandate is to help the State Department prepare for cutting edge and transformative changes, which have the potential to upend how we think about engaging with foreign publics. This report aims to achieve precisely that. In order to think carefully about public diplomacy in this ever and rapidly changing communications space, the Commission convened a group of private sector, government, and academic experts at Stanford University’s Hoover Insti¬ tution to discuss the latest research and trends in strategic communication in digital spaces. The results of that workshop, refined by a number of follow-on interviews and discussions with other organizations interested in similar questions, are included in this report. Can Public Diplomacy Survive the Internet? features essays by workshop participants that focus on emergent and potentially transformative technology and communication patterns. The essays also highlight the potential challenges and opportunities these changes create for public diplomacy practitioners in particular and the U.S. government more broadly. We explore how public diplo¬ macy practitioners can continue to productively engage with audiences around the world in the face of likely shifts in communication patterns, continue to effectively and efficiently help the United States to achieve its foreign policy priorities, and synchronize American interests with the interests of citizens and governments around the world. Respectfully Submitted, Sim Farar, Chair (California) William J. Hybl, Vice Chair (Colorado) Lyndon L. Olson, Vice Chair (Texas) Penne Korth Peacock (Texas) Anne Wedner (Illinois) Georgette Mosbacher (New York) TABLE OF CONTENTS T ransmittal Letter _ _ ii Forward: Public Diplomacy in a Post-Truth Society _ 1 Francis Fukuyama, Olivier Nomellini Senior Fellow atthe Freeman Spogli Institute for International Studies (FSI), and the Mosbacher Director of FSI’s Center on Democracy, Development, and the Rule of Law Executive Summary _ _ 2 Shawn Powers, Executive Director, U. S. Advisory Commissionon Public Diplomacy Remarks on “Public Diplomacy in a Post-Truth Society” _ 7 Bruce Wharton, Acting Under Secretary for Public Diplomacy and Public Affairs Hoover Institution, Stanford University, Stanford, California, March 20, 2017 DIGITAL’S DARK SIDE Computational Propaganda and Political Bots: An Overview _ 13 Samuel C. Woolley, Director of Research, Oxford Internet Institute’s Computational Propaganda Project Understanding the Psychology Behind Computational Propaganda _ 19 MattChessen, Foreign Service Science, Technologyand Foreign Policy FellowatThe George Washington University Rethinking Countermeasures in the Age of Computational Propaganda _ 27 Tim Hwang, Executive Director, Pacific Social Public Diplomacy’s (Misunderstood) Digital Platform Problem_ _33 Sam Ford, Research affiliate and consultant with Massachusetts Institute of Technology’s Program in Comparative Media Studies/Writing Understanding the Challenges of Artificial Intelligence and Computational Propaganda to Pub¬ lic Diplomacy _ 39 MattChessen, Foreign Service Science, Technologyand Foreign Policy FellowatThe George Washington University DISINFORMATION Psychological Principles for Public Diplomacy in an Evolving Information Ecosystem _ 49 Jeffrey T. Hancock, Professor of Communication, Stanford University Facts Matter, and People Care: An Empirical Perspective _ 55 Ethan Porter, Asst. Professor at George Washington University School of Media and Public Affairs VOA: A Weapon of Truth in the War of Words _ 61 Amanda Bennett, Director, Voice of America U.S. 201 6 Elections: A Case Study in “Inoculating” Public Opinion Against Disinformation _ 65 Jonathan Henick, Principal Deputy Coordinatorfor International Information Programs and Ryan Walsh, Senior Advisor for Digital Product, Bureau of International Information Programs In Defense of Truth, and the Threat of Disinformation _ 71 Jason Stanley, Jacob Urowsky Professorof Philosophy, Yale University NARRATIVES Public Diplomacy and Strategic Narratives _ 77 Laura J. Roselle, Professorof Political Science and International Studies, Elon University Crafting Resilient State Narratives in Post Truth Environments: Ukraine and Georgia _ 83 Vivian S. Walker, Professorof National Security Strategy, National War College America’s Strategic Narrative and a Path for Public Diplomacy _ 91 Markos Kounalakis, Visiting Fellow, Hoover Institution, Stanford University Contributor Biographies _ _ 98 FORWARD: PUBLIC DIPLOMACY I N A POST-TRUTH SOCI ETY Modern electronic communication is transforming the spread and impact of ideas in unpredictable ways. Individuals and organizations can now share in¬ formation widely and instantly at no cost, bypassing conventional media and its traditional role in curating news, focusing civic agendas, and moderating debate. While technological advances have enabled broader participation in public discussion, they have also fractured it into silos where dubious assertions and accusations can reverberate unchallenged. Fake news is disseminat¬ ed for profit or political advantage. Extremists have new forums in which to spread hatred and lies with impunity. And foreign actors influence domestic policy undetected. Compounding the problems, individuals have little ability to discern the identity of interlocutors or basis for distinguishing fact from fiction. The speed and scale of today’s “weaponization of information” is unprece¬ dented. Propelled by novelty, falsehood often travels faster than truth, leav¬ ing context and provenance behind. The traditional answer to the spread of bad information has been to inject good information into the mix, on the as¬ sumption that the truth would rise to the top. But in a world of trolls and bots, where simple facts are instantly countered by automated agents, this strat¬ egy may not be adequate. It is unclear how effectively democratic societies can continue to deliberate and function, and how hostile foreign actors can be identified and neutralized. I’m thrilled to have been supportive of the Advisory Commission on Public Di¬ plomacy’s work, in conjunction with Stanford’s Hoover Institution, on better understanding how the U.S. government can get its messages out and con¬ nect with foreign audiences in this challenging and remarkable era of global¬ ly-networked communications. Francis Fukuyama Olivier Nomellini Senior Fellow at the Freeman Spogli Institute for In¬ ternational Studies (FSI), and the Mosbacher Director of FSI’s Center on Democracy, Development, and the Rule of Law EXECUTIVE SUMMARY By Shawn Powers, Executive Director of the U.S. Advisory Commission on Public Diplomacy Scientific progress continues to accelerate, and while we’ve witnessed a revolution in communication tech¬ nologies in the past ten years, what proceeds in the next ten years may be far more transformative. It may also be extremely disruptive, challenging long held conventions behind public diplomacy (PD) programs and strategies. In order to think carefully about PD in this ever and rapidly changing communications space, the Advisory Commission on Public Diplomacy (ACPD) convened a group of private sector, government, and aca¬ demic experts at Stanford University’s Hoover Institution to discuss the latest trends in research on strategic communication in digital spaces. The results of that workshop, refined by a number of follow-on interviews and discussions, are included in this report. I encourage you to read each of the fourteen essays that follow, which are divided into three thematic sections: Digital’s Dark Side, Disinformation, and Narratives. Digital’s Dark Side focuses on the emergence of social bots, artificial intelligence, and computational propa¬ ganda. Essays in this section aim to raise awareness regarding how technology is transforming the nature of digital communication, offer ideas for competing in this space, and raise a number of important policy and research questions needing immediate attention. The Disinformation section confronts Oxford English Dic¬ tionary’s 201 6 word of the year - “post-truth” - with a series of compelling essays from practitioners, a social scientist, and philosopher on the essential roles that truth and facts play in a democratic society. Here, theory, research, and practice neatly align, suggesting it is both crucial and effective to double-down on fact-check¬ ing and evidence-based news and information programming in order to combat disinformation campaigns from our adversaries. The Narrative section concludes the report by focusing on how technology and facts are ultimately part of, and dependent on, strategic narratives. Better understanding how these narratives form, and what predicts their likely success, is necessary to think through precisely how PD can, indeed, sur¬ vive the Internet. Below are some key takeaways from the report. IN DEFENSE OF TRUTH • We are not living in a “post-truth” society. Every generation tends to think that the current generation is less honest than the previous generation. This is an old human concern, and should be seen today as a strategic narrative (see Hancock, p. 49; Roselle, p. 77). Defending the value and search for truth is crucial. As Jason Stanley notes (p. 71), “without truth, there is just power.” • Humans are remarkably bad at detecting deception. Studies show that people tend to trust what oth¬ ers say, an effect called the truth bias. This bias is actually quite rational — most of the messages that a person encounters in a day are honest, so being biased toward the truth is almost always the correct response (see Hancock, p. 49). • At the same time people are also continuously evaluating the validity of their understanding of the world. This process is called “epistemic vigilance,” a continuous process checking that the informa¬ tion that a person believes they know about the world is accurate. While we have a difficult time de¬ tecting deception from interpersonal cues, people can detect lies when they have the time, resources, and motivation. Lies are often discovered through contradicting information from a third source, or evidence that challenges a deceptive account (see Hancock, p. 49). • Fact checking can be effective, even in hyper-partisan settings (see Porter, p. 55), and is crucial for sus¬ tained democratic dialogue (Bennett, p. 61 ; Stanley, p. 71 ). Moreover, it is possible, using digital tools, to detect and effectively combat disinformation campaigns in real time (Henick and Walsh, p. 65). ACPD I Executive Summary 3 COMPUTATIONAL PROPAGANDA • Computational propaganda refers to the coordinated use of social media platforms, autonomous agents and big data directed towards the manipulation of public opinion. • Social media bots(or“web robots”) are the primary tools used in the dissemination of computational propaganda. In their most basic form, bots provide basic answers to simple questions, publish con¬ tent on a schedule ordisseminate stories in response to triggers (e.g. breaking news). Bots can have a disproportionate impact because it is easy to create a lot of them and they can post a high-volume content at a high frequency (see Woolley, p. 1 3). • Political bots aim to automate political engagement in an attempt to manipulate public opinions. They allow for massive amplification of political views and can empower a small group of people to set conversation agenda’s online. Political bots are used over social media to manufacture trends, game hashtags, megaphone particular content, spam opposition and attack journalists. The noise, spam and manipulation inherent in many bot deployment techniques threaten to disrupt civic conversations and organization worldwide (see Chessen, p. 1 9). • Advances in artificial intelligence (Al) - an evolving constellation of technologies enabling computers to simulate cognitive processes - will soon enable highly persuasive machine-generated communica¬ tions. Imagine an automated system that uses the mass of online data to infer your personality, polit¬ ical preferences, religious affiliation, demographic data and interests. It knows which news websites and social media platforms you frequent and it controls multiple user accounts on those platforms. The system dynamically creates content specifically designed to plug into your particular psychologi¬ cal frame and achieve a particular outcome (see Chessen, p. 39). • Digital tools have tremendous advantages over humans. Once an organization creates and config¬ ures a sophisticated Al bot, the marginal cost of running it on thousands or millions of user accounts is relatively low. They can operate 24/7/365 and respond to events almost immediately. Al bots can be programmed to react to certain events and create content at machine speed, shaping the narrative almost immediately. This is critical in an information environment where the first story to circulate may be the only one that people recall, even if it is untrue (see Chessen, p. 39) • PD practitioners need to consider the question of how they can create and sustain meaningful con¬ versations and engagements with audiences if the mediums typically relied upon are becoming less trusted, compromised and dominated by intelligent machines. • Challenging computational propaganda should include efforts to ensure the robustness and integ¬ rity of the marketplace of information online. Defensively, this strategy would focus on producing patterns of information exchange among groups that would make them difficult to sway using tech¬ niques of computational propaganda. Offensively, the strategy would seek to distribute the costs of counter-messaging broadly, shaping the social ecosystem to enable alternative voices to effectively challenge campaigns of misinformation (see Hwang, p. 27). In the persuasive landscape formed by social media and computational propaganda, it may be at times more effective to build tools, rather than construct a specific message. • Practitioners are not alone in their concern about the escalating use of social bots by adversarial stateactors. The privatesector is, too. Social media platformssee thistrend as a potentially existential threatto their business models, especially if the rise of bots and computational propaganda weakens users’ trust in the integrity of the platforms themselves. Coordination with private sector is key, as theirpoliciesgoverningautonomousbotswilladaptand,thus,shapewhatisandisn’tfeasibleonline. 4 ACPD I Can Public Diplomacy Survive the Internet? MOVING PAST “FOLK THEORIES” • Folk theories, or how people think a particular process works, are driving far too many digital strate¬ gies. One example of a folk theory is in the prevalence of echo chambers online, or the idea that people are increasingly digitally walled off from one another, engaging only with content that fits cognitive predispositions and preferences. • Research suggests that the more users rely on digital platforms (e.g. Twitter and Facebook) for their news and information, the more exposure they have to a multitude of sources and stories. This re¬ mains true even among partisans (though to a lesser extent than non-partisans). It turns out we hav¬ en’t digitally walled ourselves off after all (see Henick and Walsh, p. 65). • Despite increased exposure to a pluralistic media ecosystem, we are becoming more and more ideo¬ logical and partisan, and becoming more walled off at the interpersonal and physical layers. For ex¬ ample, marriages today are twice as likely to be between two people with similar political views than they were in 1960. • Understanding this gap between a robustly diverse news environment and an increasingly “siloed” physical environment is crucial to more effectively engaging with target audiences around the world. Interpersonal and in-person engagement, including exchange programs, remain crucial for effective PD moving forward (see Wharton, p. 7) • Despite this growing ideological divide, people are increasingly willing to trust one another, even com¬ plete strangers, when their goals are aligned (see the sharing economy, for example). This creates interesting opportunities for PD practitioners. Targeting strategies based on political attitudes or pro¬ files may overshadow the possibility of aligned goals on important policy and social issues (see Han¬ cock, p. 49) RETHINKING OUR DIGITAL PLATFORMS AND METRICS • Virality - the crown jewel in the social media realm - is overemphasized often at the expense of more important metrics like context and longevity. Many of the metrics used to measure the effectiveness of social media campaigns are vulnerable to manipulation, and more importantly, don’t measure en¬ gagement in any meaningful way. These metrics were built for an industry reliant on advertising for revenue generation, and as a result, may not be well-suited when applied to the context of PD (see Ford, p. 33; Woolley, p. 1 3). • Overemphasizing certain metrics, such as reach or impressions, fails to account for the risks creat¬ ed by relaying on the same portals as other, less truthful and more nefarious actors. We need to be cautious and aware of the various ways in which the digital media business industries are shaping PD content, be aware of the risks, and think carefully about safeguarding the credibility U.S. Department of State PD programs operating in this space (see Wharton, p. 7; Ford, p. 33). STRATEGIC NARRATIVES • Strategic narratives — a means for political actors to construct a shared meaning of the past, present and future of politics in order to shape the behavior of other actors.” They provide the ideological backdrop for how audiences assess the meaning and significance of current events and breaking news. Put another way, they help people make sense of what would otherwise be a dizzying onslaught of news they are exposed toon adaily basis (see Roselle, p. 77; Kounalakis, p.91). ACPD I Executive Summary 5 • Crafting effective narratives require a genuine consensus-even if limited or temporary-on our policy priorities and their underlying values, as well as a detailed understanding and appreciation of local grievances and concerns about the related policy issue (see. Wharton, p. 7; Roselle. P. 77). As such, effective strategic narratives must be mutually constructed. • Rather than focusing on trending news topics and stories alone, we need to develop greater capacity to understand competing public narratives in foreign contexts and track how they adapt over time. Understanding distinctions between system (or governance), value, and identity narratives would al¬ low PD practitioners to construct policy narratives that speak to, or at least acknowledge, the under¬ lying pillars of belief in a given community (see Walker, p. 83; Roselle, p. 77). • Every new administration creates new opportunities forforeign engagement. A shift towards a more transactionalapproachtoPD,focusedlessonvaluesbutmoreonshared policy priorities, could allow for improved relations and cooperation with a numberof countries previously hostile to American PD efforts and programs (see Kounalakis, p. 91 ). REMARKS ON “PUBLIC DIPLOMACY I N A POST-TRUTH SOC I ET Y” By Bruce Wharton, Acting Under Secretary for Public Diplomacy and Public Affairs Hoover Institution, Stanford University Stanford, California March 20, 201 7 NearlysixdecadesagoHerbertHooversaidhewant ed the Hoover Institution “to sustain for America the safeguards of the American way of life, so it is fit ting that we have gathered here today to debate a pressing challenge for both our nation and the world community the idea of a Post Truth Society, lam grateful to the Hoover Institution and to the U.S. Ad visory Commission forPublicDiplomacyforbringing ustogetherforthisimportantdiscussion. “POST TRUTH” SOCIETY There has been much discussion in the media, aca demia, and within the U.S. government about living in a post truth or post factual society and how to operate in it. Much was made of Oxford Dictionary s decision to make post truth the Word of the Year in 2016, an adjective they defined as “relating to or denoting circumstances in which objective facts are less influential in shaping public opinion than appeals to emotion and personal belief.” In such a world, the public policy debate is framed largely by what “feels true and what correlates with peoples preexisting set of beliefs and prejudices, which can often be disconnected from actual facts and the specifics of policy. It isn t so much that facts are dismissed entirely, but rather they are of second ary importance or simply not as compelling, especial ly when they challenge what feels true at an instinctu al level. In this context, all opinions have equal weight, regardless of how extreme they may be. • While this is not a new concept it has played a role in politics since antiquity in our age, social media has exacerbated the problem, accelerating the speed at which false stories spread, creating digital wild fires of misinformation. By the time a false story is out there, it is often too late to mount an effective re buttal based on facts. Compounding the problem is the active work of non state and state actors who aim not only to dissemi 8 ACPD I Can Public Diplomacy Survive the Internet? nate misinformation but, most damaging, to erode trust in traditional sources of information. These ac¬ tors — whom Get Smart fans might collectively call “KAOS” — do not necessarily want people to believe they are telling the truth, but rather to think that no one is. Theirgoal is to diminish publictrust in govern¬ ment institutions, established media outlets, and sub¬ ject matter experts, leaving citizens open to the in¬ fluence of an onslaught of questionable information generated through re-enforcing social media loops. While there is much that is accurate about this de¬ scription, I would like to contest the view that we are living in a “post-truth” society — if by that we mean truth and facts no longer matter. Facts do exist. They are out there; we cannot operate without them . And they remain compelling when they are part of a larger truth-based narrative that is backed up by support¬ ing actions. Crafting and effectively putting forth that narrative with foreign publics is the real challenge of Public Diplomacy today. Making sure “our actions match our words” is everyone’s challenge. COMPETITION FROM PSEUDO-FACTS As I said, I don’t think we are in a world beyond facts. What we are facing now is intense competition at all levels. Facts compete with pseudo-facts on sub¬ stance, on speed, and for audiences’ attention. And yes, people accept stories that “feel” true more read¬ ily than stories that challenge their beliefs. But they accept them because they believe they are true. increased since a similar poll in 2014, and both peo¬ ple who voted to leave and to remain in the EU shared much the same view. On this side of the Atlantic, polling also shows that Americans hunger for factual truth. According to a study by the Media Insight Project, a partnership of the American Press Institute and the AP-NORC Cen¬ ter for Public Affairs Research, nearly 90 percent of Americans say it is “extremely” or “very important” that the media get its facts correct. Furthermore, about 40 percent say they can remember a specific incident that eroded their confidence in the media, most often one involving inaccuracies or a perception of one-sidedness, making factual accuracy the most important component of public trust in journalism. There are also dangers in accepting a post-truth par¬ adigm. Communicators, experts, and officials may feel overwhelmed and succumb to inaction or, worse, be seduced into adopting “post-truth techniques” that appeal only to emotion and sideline facts or chal¬ lenging audiences’ beliefs. There is also the temptation to counter the barrage of misinformation by attempting to rebut every false story, but this is a losing proposition. There are too many of them, they spread too quickly, and there are too few of us to chase them. A paper published by RAND in 2016, titled “The Rus¬ sian ‘Firehose of Falsehood’ Propaganda Model,” made three important observations: 1 ) people tend to believe something when it is repeated, 2) propagan- “The way to counter pseudo-facts and misinforma¬ tion is to present a compelling narrative of our own” Brexit is often cited as an example of the post-truth phenomenon — with a leading pro-exit member of Parliament famously saying that “people in this coun¬ try have had enough of experts.” But about the same time, the Institute for Government, a British govern¬ ment organization, released a poll conducted by the research firm Populus, indicating that 85 percent of those surveyed wanted politicians to consult profes¬ sionals and experts when making difficult decisions and 83 percent wanted government to make deci¬ sions based on objective evidence. In the UK, trust in experts and confidence in government have both dists gain the advantage when they get to make the first impression, and 3) subsequent rebuttals may ac¬ tually work to reinforce the original misinformation, rather than dissipate it. The paper’s conclusion is that the most effective way to respond to misinformation is not to counter every false story out there, but to direct a “stream” of accurate messaging at whatev¬ er the firehose of falsehoods is aimed, in an effort to lead the targeted audience in a more productive di¬ rection. I agree with this approach and havesoforyears. The waytocounterpseudo-factsandmisinformationisto ACPD | Remarks on “Public Diplomacy in a Post-Truth Society” 9 present a compelling narrative of our own, one that is true, defensible, and based on the enduring values and goals that people share, not the least of which is strengthening our collective security and prosperity. To gain credibility and make our narrative relevant, we must also listen to and acknowledge our audienc¬ es’ underlying fears, grievances, and beliefs. But it is not just a matter of telling a good story; the narrative must be tied to action. A case in point is the history of space exploration in this country, in particular the quest to put a man on the moon. In the Cold War context, this effort was an important security goal, one that required public sup¬ port, resources, and full political commitment over many years. In 1961, President Kennedy gave his his¬ toric speech before a joint session of Congress that set the United States on a course to the moon, which he followed with other speeches and public acts that inspired not just the American people, but invited au¬ diences around the globe to be part of this great en¬ deavor. And foreign publics responded by embracing U.S. aspirations on behalf of the human race. When the entire planet watched Neil Armstrong alight from the Eagle lunar module and utter the phrase “That’s one small step for man, one giant leap for mankind,” he distilled into these few words a decade-long nar¬ rative that fueled the imagination and hopes of bil¬ lions around the world. Another example is the Voice of America. I am glad Amanda Bennet is here and will talk more about this later, but something she said last week at a public meeting on the Hill really struck me. She described VOA as “exporting the First Amendment” - that is to say, the value and importance that Americans place on a free and independent press - by providing fact- based, balanced reporting to millions of people in closedsocietieseveryday.Wetalkaboutit.andwedo it. Nothing is as powerful as a living example, and as a Public Diplomacy practitioner, I could not be prouder of the message, both literal and figurative, that VOA delivers. In short, we’ve got to “walk the talk,” or risk losing credibility. This is not to say countering disinforma¬ tion is easy. It requires strategic thought, creative tactics, and sustained investment. The State Depart¬ ment and other parts of the federal government have been focused on this issue for several years, and analyzing how these efforts have fared is helping us chart the way ahead. CASE STUDY - STATE’S APPROACH TO FIGHTING EXTREMIST IDEOLOGY After the 9/1 1 terrorist attacks, we in government — and those of you in academia and the thinktank world — were desperate to find explanations for what had happened and, more importantly, to prevent some¬ thing similar from happening again, with a particular focus on containing and countering the appeal of vio¬ lent extremist ideology. All ideas were encouraged, and we pressed our peo¬ ple to think creatively and to try new approaches. One approach aimed at mass appeal was the $15 million “Shared Values” campaign featuring Muslims living happily in the United States. As well intended as this was, the messaging did not acknowledge under¬ lying grievances and was not considered effective in reaching young Muslim audiences overseas. Another idea you may remember from a just few years ago was the “Welcome to ISIS Land” video, which went viral for all the wrong reasons. It was heavily criticized for embracing the enemy’s tac¬ tics and coming across as bullying. Most critically, it proved to be ineffective as the U.S. government was not a credible source of information for the intend¬ ed audience, who only seemed to be alienated by the message. Hampering our efforts was an inability to measure the impact of our work reliably. For instance, the former Center for Strategic Counterterrorism Com¬ munications (CSCC), which was established in 2010 to counter extremist ideology, could point to the size of its Facebook and Twitter followings — and the number of death threats and efforts to shut down its accounts were evidence that the center had gotten undertheskin of ISIS — but it could not measure ef¬ fectiveness. As a result, itwasneverclearwhetherits efforts reached those at risk of joining ISIS, let alone diverted them from that path. The CSCC was also under resourced. Its budget hov¬ ered in the range of $5-6 million per year, while the Pentagon was spending about $150 million on simi¬ lar efforts and the CIA even more. This situation even emerged as a media story, with ABC News describing the U.S. government’s messaging strategy to counter extremist ideology as underfunded and ineffective. Thisexperienceprovideduswithawealthofvaluable lessonsforcharting a newwayforward in countering 10 ACPD I Can Public Diplomacy Survive the Internet? false narratives, including: • Not imitating the enemy, • Having a credible message based on facts and evidence that acknowledge underlying grievances, • Partnering with credible, independent, trust¬ ed messengers, • Using technology to identify the right audi¬ ences and the best approaches for reaching them, • Employing analytics to evaluate effectiveness and feeding that information back into the process, and • Securing political and bureaucratic support, including sufficient funding and personnel. On the technology front, I am particularly enthusi¬ astic about the potential to use tools such as social graph analysis (SGA) to help us identify credible indi¬ viduals who drive and shape online opinion within e ach country. Network analysis can provide informa¬ tion in two critical areas: 1 ) topics important to people in target audiences and 2) the most uniquely influen¬ tial people within those topical clusters. This informa¬ tion, which is used daily by business to analyze con¬ sumers’ tastes and persuade them to buy more, can provide a clearer view for engaging target audiences in partnership with the influencers they trust most. We in the U.S. government are prohibited from using such tools when the information of U.S. citizens is in¬ volved. ANEW APPROACH The beneficiary of these lessons is the State Depart¬ ment’s new Global Engagement Center (GEC), which is legislatively given the task “to lead, synchronize, and coordinate efforts of the Federal Government to recognize, understand, expose, and counter foreign state and non-state propaganda and disinformation efforts aimed at undermining United States nation¬ al security interests.” In this role, the GEC leads the interagency in developing a whole-of-government approach to countering malign actors in the informa¬ tion space, seeking to fully leverage the strengths and capabilities of each agency involved in this effort. A key element to ensuring coordination and maximum efficacy is an interagency synchronization meeting hostedweeklybyGEC. The GEC also enjoys strong support on the Hill, from both sides of the aisle. In fact, it was Congress that expanded the GEC’s mandate — which originally fo¬ cused solely on non-state actors — to include recog¬ nizing, understanding, and exposing state-sponsored propaganda and disinformation and countering its effects. In terms of resources, the GEC is funded at approx¬ imately $16 million dollars for FY-17 and is slated to have an additional $19.8 million in supplemental funding in FY-18. Further, Congress has authorized - although not mandated - the Department of Defense to transfer up to $60 million a year, in both FY-17 and FY-1 8,tosupportGECactivities. We are focusing today on the importance of facts, and central to the work of the GEC is injecting fac¬ tual content into the information space to counter violent extremist radicalization and recruitment. Content is developed through collaborative, thematic campaigns in coordination with the U.S. interagency and with members of the Counter-ISIS Coalition and other global partners. GEC support includes funding, technical assistance, capacity building, and conceiv¬ ing and implementing joint projects. Using this approach, we have reduced direct engage¬ ment on violent extremism in favor of partner-driv¬ en messaging at the local level. These partners are credible voices that can deliver messages that reso¬ nate with at-risk populations, such as NGOs, schools, young people, social and civil society leaders, reli¬ gious leaders, and governments. Additionally, the GEC is utilizing data science from both the public and private sectors — including polling operations, audience studies, and academic research — to identify and understand target audi¬ ences, to guide and inform the development of mes¬ saging and content, and to measure effectiveness. For instance, the GEC’s “Defectors” campaign used content from 14 Coalition countries that highlighted the lived experiences of ISIS defectors and the ef¬ fects of their recruitment on theirfamilies. In just one week, the campaign reached 2.4 million people who watched over one million minutes of video. Ultimate¬ ly, the Defectors campaign reached seven million individuals and garnered 780,000 “click-throughs” from people identified as being at risk for recruit¬ ment by violent extremists. Despite the impressive ACPD | Remarks on “Public Diplomacy in a Post-Truth Society” 11 numbers, the cost of this data-driven campaign was only$15,000. Of course, the GEC is still fairly new, so I look forward to letting you know in future discussions how it is far¬ ing. But I think we are on the right track in countering an ideology thattrades in falsehoods by working with credible partners to present the facts and alterna¬ tives that are true. CONCLUSION Going back to my original premise, I respectful¬ ly disagree with the concept that we are living in a “post-truth society.” What we are facing instead is in¬ creased competition from pseudo-facts, but the truth is still valued, desired, and ultimately compelling. We just need to find the right ways to communicate it. And while some of my remarks have been focused on the messaging component of Public Diplomacy, we must remember that many other PD tools play chaff are vital. One way we are doing this is through TechCamps focused on disinformation. These inter¬ active workshops, led by technical experts, build the capacity of key foreign influencers in civil society to push back on fake news. A special fund dedicated to incubating collaborative follow-on projects maximiz¬ es each workshop’s impact and has resulted in such innovations as a one-stop data verification tool for Ukrainian journalists to fact-check online media con¬ tent. To be truly effective, however, we must start at a younger age. A recent study by Stanford showed that students at most grade levels cannot tell the difference between fake and real news as they often lack the critical thinking skills needed to separate truth from misinformation. Game theory has the po¬ tential to help us develop smarter ways to build the fact-checking skills of students, and video games could contain elements that help players of all ages become more aware - and wary of - faux facts. This realization has prompted some teachers across the “Our challenges are too big and our resources too limited to go it alone. ” a vital role in sharing the truth, such as educational and cultural exchanges, youth initiatives, and English teaching programs. These types of people-to-people interactions help reframe conversations on conten¬ tious issues, demonstrate the value of transparency, and build trust with key audiences. Finally, thereisonelastcriticalelementinthisdebate. In addition to offering compelling, truthful narratives, I believe we must also help foreign audiences target¬ ed by concerted disinformation campaigns to better understand the dangers of accepting everything at face valueand encourage them tocultivatea“healthy skepticism.” By this I do not mean to promote para¬ noia, simply vigilance. But how do we do this effec¬ tively when people, especially young people, are bom¬ barded with so much dubious information? How do we help them become healthy skeptics? Training and education programs that both cultivate a questioning mindset and build the skills of infor¬ mation consumers to separate the wheat from the country to use games, such as Simon says, to help students build these skills. Beyond these ideas, I believe we should be asking what economic mechanisms might be used to en¬ courage skepticism and objective truths. Are there known business models that reward honesty and pe¬ nalize dishonesty? Perhaps some of you here may be tempted to undertake research in these areas. I look forward to discussing these issues with you fur¬ ther, but before we open the floor for questions and comments, I want to thank all of you for your interest in, and support for, the work of Public Diplomacy. We in government need your input, your ideas, and your talent. Our challenges are too big and our resources too I i m ited to go it alone. T ogether, I know we ca n suc- cessfully navigate the current sea of misinformation and propaganda and find a productive path forward. Afterall, we have truth on ourside. COMPUTATIONAL PROPAGANDA AND POLITICAL BOTS: AN OVERVIEW By Samuel C. Woolley, Director of Research, Oxford Internet Institute’s Computational Propaganda Project Computational propaganda is best defined as the assemblage of social media platforms, autonomous agents and big data directed towards the manipu¬ lation of public opinion. Social media bots are the primary tools used in the dissemination of computa¬ tional propaganda. When bots are used to automate political engagement in attempts to manipulate pub¬ lic opinion our team at the University of Oxford calls them “political” bots. Political bots allow for massive amplification of political views, they can empower a small group of people to set conversation agenda’s online. They are used over social media to manufac¬ ture trends, game hashtags, megaphone particu¬ lar content, spam opposition and attack journalists. When in the hands of powerful, well-resourced, polit¬ ical actors these automated tools can be used to both boost and silence communication and organization among citizens in both democratic and authoritarian regimes. Security experts argue that more than 1 0 percent of content across social media websites, and 62 per¬ cent of all web traffic, is generated by bots — pieces of computer code that automate human tasks online. Bots dominate many mundane tasks on the internet, from aiding in the generation of personal online news preferences, to ad generation, to promoting matches on social media platforms to undertake tasks and mimic real users. Over the last four years, numerous news outlets, from The New York Times to The Guardian, have covered rising and evolving usage of bots. They attempt to explain how these socially oriented au¬ tomated scripts work in specific contexts, from the world of online dating to that of real-time ad shar¬ ing. The ways bots are being deployed, however, are evolving beyond social spheres to those discretely political. Politicians, governments and military or¬ ganizations have begun using a special variety of bot software to manipulate political communication and engagement, choke off debate and muddy political issues. Until roughly five years ago, social bots were most¬ ly harnessed by technologically adept marketers to send spam in the form of automatically tweeted ad¬ vertising content. Politicians have taken note of and emulated celebrity Twitter users’ tactics of purchas¬ ing massive amounts of bots to significantly boost follower numbers. Militaries, state-contracted firms and elected officials now use political bots to inva- sively spread various forms of propaganda and flood newsfeeds with political spam. Recent research re¬ veals the pervasive breadth of global political bot use across online social networks. For instance, re- “ Security experts argue that more than 10 percent of content across social media web¬ sites, and 62 percent of all web traffic, is gen¬ erated by bots” on social media. News organizations now use bots to track and disseminate breaking articles. Sites like Wikipedia, which generate publically accessible knowledge, use bots as an essential part of their la¬ bor force. The latest social bots are automated software used searchers at the University of Southern California believe as many as 48 million accounts (around 15 percent) on Twitter are actually bots. This number is a 7 percent increase from the projection of automat¬ ed accounts that Twitter gave in an 2014 SEC report. Moreover, bots have been the main tools for online astroturf1 and smear campaigns during political mo- 14 ACPD I Can Public Diplomacy Survive the Internet? ments worldwide: the U.S. midterm elections of 2010, the ongoing crisis in Syria, and the 2014-15 disputes over Crimea. Politically oriented bots are an emergent phenomena and are amongst the most important recent innova¬ tions in political strategy and communication tech¬ nology. Bots are prevalent and active in social media conversations — and their presence in these spaces continues to grow. The noise, spam and manipulation inherent in many bot deployment techniques threat¬ en to disrupt civic conversations and organization worldwide. QUESTIONS Several questions are at the heart of research on the automated tools known as “political” bots and on their chief output, computational propaganda: • How does the use of this technology affect public opinion, or behavior around voting and civic en¬ gagement? • What do internet-oriented companies, particu¬ larly social networking platforms like Twitter and Facebook, do to track and curate political content generated by bots? • How are bots used by or against other demo¬ cratic institutions, particularly the free press and non-governmental organizations, to generate or influence content and communication? • How are bots challenging traditional notions of agency in the field of science and technology studies and traditional conceptualizations of “the actor” as a unit of study in political communica¬ tion? UNDERSTANDING COMPUTATIONAL PROPAGANDA Relatively little academic work — especially empirical research focusing on critical social considerations — has been done on social bots and the processes associated with them, within the social sciences. Political automation, especially automation that chal¬ lenges the foundations of civic life, is a significant techno-cultural phenomenon. It is also one that was unforeseen by the early 2000s social platform move¬ ment. The social ties of those who work within the automation work must be studied to build knowledge on the larger industries and organizations that dom¬ inate the digital sphere. To that end, more academic work must be done to build understandings via first¬ hand interaction with the people who build and de¬ ploy political bots. As Markham and Baym argue, field research and other qualitative methods are critical to adding breadth in understandings of emergent techno-social phenomena — particularly online. Many computer and social scientists treat bot-gen- erated traffic as a nuisance to be detected and man¬ aged, thus extant systems work to simply identify or block accounts that appear to be running as automat¬ ic scripts. This approach is too simplistic and avoids focusing on the larger, systemic problems presented by political bot software. Political bots suppress free expression and civic innovation via the demobiliza¬ tion of activist groups and the suffocation of demo¬ cratic free speech. Political bots and computational propaganda must, therefore, be better understood for the sake of free speech and digitally mediated civic engagement. The information that exists on po¬ litical bots is disjointed and often isolated to specific, country or election-oriented, events. BOT TYPES AND CASES OF USE Any taxonomy of social bots should begin by discuss¬ ing the tasks for which public facing bots are built. Within this discussion are questions about the ca¬ pabilities and sophistication of these automated and semi-automated social actors. A simple typology of bots should be based upon technical input and com¬ municative output. This method of distinction runs on a scale from simple bots preprogrammed with a corpus of simple phrases or words that are then transmitted to internet viewers to smart bots using machine learning tactics to scrape data from both websites and crowd-sourced interactions to then communicate with publics in a unique and, potential¬ ly, unforeseen way. An example of a simple bot in the former category could be a spam bot that sends out the same com¬ mercial link to users on a particular platform or an art-oriented bot like @everyword that tweets all words in the English language from the Oxford En¬ glish Dictionary overasetperiodofyears. Microsoft’s Tay is an example of a smart bot — what developers, and indeed Microsoft itself, deem an “Al chat bot.” Tay, regardless of the public relations nightmare she caused for Microsoft, was built to learn from those ACPD I Computational Propaganda and Political Bots: An Overview 15 around her. Her communicative and computational functionality was somewhat rudimentary — she could be prompted to reiterate phrases by simply being told “Tay, repeat after me...” — but this machine learning, crowd-sourcing, net-scraping bot is indicative of the direction of similar social assistance/communication software: Cortana, Siri, Viv, Google Now, Alexa and others. Another way of envisioning a typology of social bots is through questions of transparency. Simply put, is the bot transparently a bot or does it pretend to be human? The last five years have seen a rise in the so¬ cio-political use of bots that pass themselves off as human — both implicitly through a lack of bot identify¬ ing information and explicitly by claiming outright to be human — in attempts to manipulate public opinion, stymie activism and perpetuate trolling of opposi¬ tion. There are three general types of transparency when it comes to bots: 1 ) transparent bots — bots that are clearly labeled as bots (e.g., @sortingbot, which sorts Twitter users into Hogwarts houses); 2) semi¬ transparent bots — those that claim to be bots, but are human-like or that have real-time dual human/ computational input (e.g., Facebook’s “M,” which can answer questions through automated computational search or by accessing an in-house human team; or Tay, which is a bot that acts like a teenage girl); and 3) nontransparent bots — bots that claim to be human (e.g., the bots used by the Mexican government in at¬ tempts to sway public opinion during the 201 2 presi¬ dential election). A typology of political bots can be built that usefully catalogues the most pervasive uses of bot technology and reveals emergent patterns may be helpful. Gov- Timeline of major developments in bots and their political use 2010 (a l lorse ebooks, allegedly a twitter bot, becomes an internet phenomenon Journalists at Gawker reveal that [a Horse ebooks is human-driven ******* Mexican government officials begin using twit¬ ter bots to silence dissent. Opponents ••••••• respond with their own bot -driven attacks Twitter reveals to SEC that 1 2% of accounts on •••••• the platform are bots Humor bot (a oliviataters is launched by comedian Ron Dubbin Reddit Moderator Bots on the /technology ••••••• subrcddit allegedly delete comments related to the Snowden revelations and the US National Security Administration Microsoft's Twitterbot “Tay," billed as an Al chatbot, is fooled into publicly tweeting racist, misogynist, and generally offensive content ••••••• CEO of messanger app Kik predicts a “bot •••••• goldrush" 2011 Supporters of Republican Senate candidate Scott Brown use Twitter Bots to attack, and spread propaganda, about opponent Martha Coakley 2012 • • • • • Google Now. a digital personal assistant, is launched US Presidential candidate Mitt Romney is accused of using bots to boost his Twitter following 2014 • • • • • Bot Eugene Goostman passes the Turing test ••••• Turkish Prime Minister Recep Tayyip Erdogan allegedly sanctions thousands of bots to support his party and attack opponents 2015 • • • • • Coca Cola’s bot-driven twitter account tweets sections of “Mein Kampf' 2016 2016 Facebook launches a bot platform for Messanger • •••• US Presidential Candidate Donald Trump is accused of using bots to attack opponents and boost online credibility 2017 Image credit: Samantha Shorey 16 ACPD I Can Public Diplomacy Survive the Internet? ernments and other political actors most generally deployed political bots during elections or moments of distinct, and country-specific, political conversa¬ tion or crisis. For example, bots used in Venezuela focused solely on attempts to manipulate public opin¬ ion in state. The Syrian government has reportedly used bots to generate pro-regime propaganda tar¬ geted at both in state and external targets on Twitter during the ongoing revolution. In both these cases, bots were created in response to the local political climate. Though the ways in which political bots have been used varies from country to country and political in¬ stance to political instance, there are three primary types of political bots: 1 ) follower bots — those used to boost political figures’ follower numbers and passive¬ ly like or re-tweet content; 2) roadblock bots — those used to spam hashtags associated with activists or political opposition in order to shut down or interrupt dissent via non-traditional communication channels; and 3) propaganda bots — those used to mimic hu¬ mans while sending out effusively positive informa¬ tion about an embattled government or politician or to propagate negative attacks against the opposition. Follower Bots: Follower bots have also been used during elections and security crises to pad politicians’ social media follower lists. In these cases, politicians buy bot followers — which mimic real human users — in attempts to look more politically relevant or tech¬ nologically savvy. There are several prominent ex¬ amples, particularly in Western states. According to Inside Croydon, UK political candidate Lee Jasper used bots to boost the number of his Twitter followers in order “to give a false impression of the popularity of his campaign.” There was a similar bid by former U.S. presidential candidate Mitt Romney in which political bots were used for padding his social media follow¬ ers. According to an NBC article, “[in] over 24 hours starting July 21, the presumptive Republican nomi¬ nee acquired nearly 117,000 followers — an increase of about 17 percent.” This increase is so substantial it is unlikely to have occurred naturally, through the actions of human twitter users. Roadblock Bots: During elections, roadblock bots have been used to demobilize an opposing party’s followers. In this case, the deployer sends out Twit¬ ter “bombs:” barrages of tweets from a multitude of bot-driven accounts. These tweets co-opt tags com¬ monly used by supporters of the opposing party and re-tweet them thousands of times in an attempt to prevent detractors from organizing. For instance, if a political actor notices that their opponent’s support¬ ers consistently use the tag #freedomofspeech in or¬ ganizational messages, then that actor might make an army of bots to prolifically re-tweet this specific tag. The effect of this is that the opponent’s support¬ ers have a very difficult time searching common tags in attempts to organize and communicate with their fellows. Propaganda Bots: Many cases of propaganda bot use occurwhen governments target perceived cyber-se¬ curity threats or political-cultural threats from other states. The mostwidely reported case of state-sanc¬ tioned propaganda bots occurred in Russia. In this instance, Russian bots were allegedly used to pro¬ mote regime ideals or combat anti-regime speech against targets abroad. Chinese propaganda bots have also attacked other countries and commercial entities. Political actors in Azerbaijan, Iran and Mo¬ rocco reportedly used propaganda bots in attempts to combat anti-regime speech and promote the ide¬ als of the state. Governments, politicians and contractors employ combinations of both propaganda and roadblock bots to attack in-state targets on social media. De¬ scriptions of bot usage in Mexico are particularly representative of this automated strategy. Accord¬ ing to numerous sources, the Mexican government has used Twitter bot armies to stifle public dissent and effectively silence opposition through spam tac¬ tics. Penabots, named after the Mexican President Enrique Pena Nieto, have also been used to send out pro-government propaganda. In Turkey, journalists report that both President Recep Tayyip Erdogan’s government and the opposition Republican People’s Army have used political bots against one another in efforts to spread propaganda, fight criticism and block dissent. In China, and in the Chinese administrative regions of Tibet and Taiwan, bots have been used to quash sov¬ ereignty movements while promoting state ideals. According to journalist Brian Krebs, “Tibetan sym¬ pathizers [...] noticed that several Twitter hashtags related to the conflict — including #tibet and #freet- ibet — are now so constantly inundated with junk tweets from apparently automated Twitter accounts that the hashtags have ceased to become a useful way to track the conflict.” Propaganda bots have been used during elections to send out pro-government or pro-candidate social ACPD I Computational Propaganda and Political Bots: An Overview 17 media messages. The New York Times points to South Korean state prosecutors’ allegations that “agents from the National Intelligence Service of South Ko¬ rea posted more than 1 .2 million Twitter messages last year to try to sway public opinion in favor of Park Geun-hye, then a presidential candidate, and her party ahead of elections in 201 2.” Geun-hye eventu¬ ally won the presidency, but the intelligence chief in charge of the bot-driven effort was jailed and re¬ mains in prison. Geun-hye has since been ousted and indicted on charges in ofwidespread corruption. Our team at Oxford has also found that political bots — automated accounts used over social media to spread political content — had a significant role in spreading information and misinformation during the U.K.’s Brexit referendum. Our report on Twitter and Brexit found that the family of hashtags associ¬ ated with the argument for leaving the EU dominated both general conversation. Parts of this conversation were driven by armies of political bots messaging ef¬ fusively in support of the “leave” side. Furthermore, we found that less than 1 percent of sampled accounts — most of them highly automated — generated almost one-third of all messages about the leave campaign. More work, however, needs to be done to determine whether the use of political bots had a specific effect upon electoral outcomes: did automated messaging change the way people voted? CONCLUSION The study of computational propaganda is, by its very nature, a transdisciplinary endeavor. Such work necessitates a combination of disciplinary and pro¬ fessional input from fields ranging from psycholo¬ gy to information science, public policy to machine learning, political science to sociology. The spread of political bots, and associated content from mis¬ information to intimidation, complicates the ways in which politics are conducted and perceived both on and offline. While tools, such as BotorNot and Twitter Audit, are able to determine an account or users au¬ tomation levels, they are not equipped for doing the deeper work of uncovering coordinated attacks or botnets. No effective tools currently exist for uncov¬ ering who proliferates political bot attacks or where attacks originate. There is also a need for basic un¬ derstandings of how the use of political bots plays out in both local and comparative contexts. Our project at Oxford is currently working on a series of case stud¬ ies, to be released in mid-June 2017, that will analyze the role of bots across several countries. This report will provide key insight into the ways this technology is being developed and deployed, but also regulated and resisted, in Europe, North America, South Amer¬ ica, Africa and Asia. More work needs to be done to study the effects of computational propaganda. Though social me¬ dia platforms, and tools like political bots, are being harnessed to spread misinformation in attempts to manipulate public opinion, little is known about how their use changes actual voting behavior or more flu¬ id aspects of public life such as citizens’ conceptions of a particularcandidate or issue. Global policy mak¬ ers, academics, technology professionals and others must work together to build continued understand¬ ings of this rapidly progressing phenomenon. Moreover, these same actors and their communities must begin to generate solutions to the problems of computational propaganda. Hypothetical fixes, in¬ cluding both defensive and offensive measures, must transcend pure technological or social solutionism and combine both software tools and media litera¬ cy — and other, yet unforeseen and unconsidered ap¬ proaches and efforts — in order to succeed. ENDNOTES 1. Astroturfing refers to the practice of masking the sponsors of a message or organization (e.g., political, advertising, religious or public relations) to make it appear as though it originates from and is supported by a grassroots participant(s). UNDERSTANDING THE PSYCHOLOGY BEHIND COMPUTATIONAL PROPAGANDA By Matt Chessen, Foreign Service Science, Technology and Foreign Policy Fellow at The George Washington University Machine-driven communications tools (MADCOMs) are frequently used by a variety of actors to spread ideas online. Computational propagandists increas¬ ingly use these tools for influence and disinformation. Their effectiveness is based on principles from cog¬ nitive psychology and the science of persuasion. This paper is a companion piece to the accompanying articles on computational propaganda, MADCOMs and artificial intelligence tools and will illustrate how these technologies exploit persuasive techniques. MACHINEDRIVENCOMMUNICATION TOOLS (MADCOMS) The most commonly used MADCOMs are simple bots (web robots) that post content on social media, web¬ sites, comment sections and the like.1 Their current capabilities are limited to providing basic answers to simple questions, publishing content on a sched¬ ule or disseminating content in response to triggers. However, bots can have a disproportionate impact because it is easy to create a lot of them and bots can post content with high volume and high frequency. Little expertise is required to run simple bots. An in¬ dividual can easily operate hundreds of Twitter bots with minor technical knowledge using readily avail¬ able hardware and software. Bots and other MAD¬ COMs are currently used by corporations, politicians, hackers, individuals, state-sponsored groups, NGOs and terrorist organizations in an effort to influence conversations online. Bot users’ goals are myriad: • Individuals use MADCOMs for many pur¬ poses, including making profits, making the world a better place or making mischief. • Academics use MADCOMs to network with their communities, share ideas and conduct research. • Organizations use MADCOMs to gain support for their causes, inform a wider range of peo¬ ple, and connect disparate and dispersed ac¬ tivist groups. • Companies use MADCOMs for marketing, persuading you to purchase their product or service. They also use MADCOMs for custom¬ er service and as human-like “faces” for fulfill¬ ing back-end business processes. • Politicians use MADCOMS to create the ap¬ pearance of massive grassroots support (astroturfing), to amplify messages and sup¬ press opposition communications. • Terrorist and hate groups use MADCOMs to spread their messages of intolerance, to sup¬ press opposition efforts and to identify new recruits. • Nations use MADCOMs for public diplomacy, service delivery, propaganda, counter-mes¬ saging, disinformation, espionage, democ¬ racy suppression and intimidation. In the future, networks of competing, state-spon¬ sored artificial intelligence MADCOMs will use human-like speech to dominate the in¬ formation-space and capture the attention of the most online users. In short, all groups may use MADCOMs for political purposes, including persuasion, disinformation, as¬ troturfing, undermining speech, intimidation, doxing and distraction from politically inconvenient topics. A SIMPLE TAXONOMY OF MACHINE-DRIVEN DISINFORMATION AND PROPAGANDA Computational propaganda is a new term for the use of machine-driven communication tools and as¬ sociated technologies for political purposes. These purposes can range from relatively benign amplifi¬ cation of political messages to insidious state-spon¬ sored trolling and disinformation. Computational propaganda typically uses simple bots to influence conversations online. These bots operate on social media user accounts that may have sophisticated, human-like profiles. 20 ACPD I Can Public Diplomacy Survive the Internet? “ Security experts argue that more than 10 percent of content across social media web¬ sites, and 62 percent of all web traffic, is gen¬ erated by bots” Bots typically follow three general patterns of behav¬ ior: • Propaganda bots attempt to persuade and in¬ fluence by spreading truths, half-truths and out¬ right fake news in a high volume or in response to triggers (e.g., keywords or a politician’s tweet). • Follower bots fake the appearance of broad sup¬ port for an idea or person. They can hijack algo¬ rithms that determine trending news or trending people by generating “likes” for content or by fol¬ lowing users en masse. • Roadblock bots undermine speech by diverting conversations. This could be relatively benign — like nationalist cheerleading or a “look at this cat video” type of distraction. Or it could be more in¬ sidious — like spamming hashtags used by activ¬ ists so their topical conversations and coordina¬ tion are overwhelmed with gibberish. At their most extreme, bots are used to troll/intimi¬ date journalists, activists and others into silence by bombarding them with thousands of threatening or hateful messages. (Note: for more information on computational propaganda, see the accompany¬ ing papers by Samuel Woolley, Tim Hwang and Matt Chessen). Computational propaganda techniques have also been combined with more traditional hacking meth¬ ods — like disclosures of information from politician email accounts or distributed denial of service at¬ tacks on election monitoring websites and apps — and are typically used as elements of a larger infor¬ mation strategy.2 HOW MACHINES EXPLOIT VULNERA¬ BILITIES IN HUMAN MINDS Computational propaganda has its roots in tradition¬ al propaganda, cognitive psychology and the science of persuasion. Computational propaganda tools ex¬ ploit a number of traditional theories of influence and persuasion, including: • Variety of sources: Multiple sources, prefer¬ ably presenting different arguments leading to the same conclusion, are more persuasive than single-channel, single-message cam¬ paigns.3 And the volume of different argu¬ ments supporting a conclusion are more im¬ portant than the quality of the actual individual arguments.4 Bots allow propagandists to use thousands of social media accounts to circu¬ late a high volume of messages from multiple online sources using text, images and video, all pointing to the same conclusion. Bots can outperform humans by posting content con¬ sistently throughout the day, or by spamming high-volume content in response to specific triggers. This constant repetition of themes through multiple channels creates the appear¬ ance of consensus. • Number, volume and variety of endorsements: Endorsement by large numbers of users, re¬ gardless of their individual credibility, boosts persuasiveness.5 In information rich envi¬ ronments, people favor the opinions of highly endorsed users over experts.6 Follower bots allow propagandists to generate high-volume likes and follows for selected content and us¬ ers. Propaganda bot networks will retweet and share content among machine-driven ac¬ counts, creating the perception of mass sup¬ port. This astroturfing (faking the appearance of grassroots support) can push low-quali- ACPD I Understanding the Psychology Behind Computational Propaganda 21 ty, questionable or outright false content to the top of trending topics lists, enhancing its credibility and persuasiveness. In the high-in- formation online environment, this mass user endorsement trumps expert views. • Social proof from others: The psychological theory of implicit egotism explains that hu¬ mans have an unconscious preference for things they associate with themselves.7 Re¬ cipients are more likely to believe messages from users they perceive as similar to them¬ selves. People believe sources are credible if they think other people believe them credible. Popular users and content are perceived as more important. Propagandists often cre¬ ate user profiles for bot accounts with imag¬ es, usernames and background information that is similar to their target audience. The audience likely doesn’t know the account is machine-driven and believes it is another hu¬ man with similar interests and demograph¬ ics. Bot-driven accounts follow real users and other bot-driven accounts en masse, creating the perception of a large following. This large following enhances perceived credibility, at¬ tracting more human followers and creating a positive feedback cycle. • The false consensus effect is a cognitive bias where people overestimate the extent to which their views reflect wider society.8 It is espe¬ cially prevalent when individuals are in groups that reinforce their beliefs. People think it’s appropriate to believe, feel or act when they think that people comparable to them are also believing, feeling or acting in the same way.9 When computational propagandists use bot networks to troll individual users or groups with hate speech, it creates the perception among some users that this is acceptable behavior. This behavior then becomes nor¬ malized among the group and human users will replicate the trolling behavior without bot provocation.10 • Mass criticism undermines expertise and trustworthiness: Mass attacks on the credi¬ bility of messengers diminishes theirtrustand credibility and reduces the chance that users will act on their content.11 Propagandists use bots for mass trolling attacks on human users (like journalists, rights activists and experts) and competing networks of users and bots that contradict their messaging operation. Propagandist attacks may present multiple al¬ ternative arguments that undermine credibil¬ ity through volume rather than quality. These may be combined with personal attacks, hate speech, trolling and doxing intended to intim¬ idate the user and frighten them into silence. False consensus effects can result in human users “piling on” believing that their communi¬ ty finds this behavior acceptable.12 • Conversion theory of minority influence: Mi- nority groups can have disproportionate influ¬ ence- over the majority by expressing a confi¬ dent, consistent message over time.13 Bots can disseminate high-volume content constantly, with significant sharing between bots, creat¬ ing the appearance of a tight-knit community with unwavering beliefs. • The authority principle: People are more like¬ ly to believe others who look like they know what they are doing or are in positions of power.14 Propagandists frequently create ma¬ chine-driven accounts with false credentials — like affiliation with government agencies, corporations, political parties, etc. — to boost credibility. • The illusory truth effect: People believe mes¬ sages to be true after repeated exposure, even if ridiculous. Familiar messages are also critiqued with less precision than unfamiliar ones.15 Propagandists generate “truthiness” by using bots to spam our feeds with high-vol- ume content supporting their ideas. Over time, these messages become familiar and more likely to be accepted. • Belief perseverance, motivated reasoning and the first-mover advantage: Once a person forms a belief it can be difficult to change his or her mind, even if the information creating the belief is patently false and factual informa¬ tion is later presented.16 In fact, corrections can actually reinforce confidence in the original misinformation.17 Corrections are especial¬ ly ineffective where the correct information threatens a person’s self-identity or world¬ view.18 Even if people understand and accept corrections, the initial disinformation can have lingering, negative, attitudinal impacts called “belief echoes.”19 Computational propaganda can shape false narratives broadly and quickly, 22 ACPD I Can Public Diplomacy Survive the Internet? making it difficult for factual, well-researched or fact-checked messages to gain traction. Even if corrected, the rapid activity of bots and virality of social networks can cause lingering belief echoes that can be reinforced over time. Opinionated human pundits generate false be¬ liefs but bot networks have greater reach and volume and are far more insidious.20 IMPLICATIONS FOR PUBLIC DIPLO¬ MACY EFFORTS Computational propaganda is not a vision of the fu¬ ture. Computational propagandists are using MAD- COMs now to exploit all of these persuasive tech¬ niques. Emerging artificial intelligence technologies will improve the effectiveness of MADCOMs and computational propaganda significantly over the next several years.21 These insights from cognitive psychology and per¬ suasion may imply or suggest best practices for pub¬ lic diplomacy professionals, but there are multiple — and sometimes conflicting — perspectives on nearly any topic. Emily Thorson, an assistant professor of media and public affairs at the George Washington University, illustrates one facet of this complexity: “The existence of belief echoes provides an enormous incentive for politicians to strategically spread false information with the goal of shaping public opinion on key issues. However, results from two more experiments show that politicians also suffer consequences for making false claims, an encouraging finding that has the potential to constrain the behavior of politicians presented with the opportunity to strategically create belief echoes. While the existence of be¬ lief echoes may also provide a disincen¬ tive for the media to engage in serious fact-checking, evidence also suggests that such efforts can also have positive consequences by increasing citizens’ trust in media”22 Similarly, much of the literature suggests that di¬ rectly counter-messaging disinformation with cor¬ rections may be ineffective or counterproductive. This implies that any counter-messaging should be focused on short-circuiting misinformation before it goes viral, a difficult and resource intensive proposi¬ tion. This messaging would be immunological rather than counter — designed to build resistance to the disinformation in targeted communities before the disinformation has time to infect them, rather than directly contradicting the disinformation. However, other research suggests that there are specific con¬ ditions under which corrections of disinformation can be effective.23 This would call for more selec¬ tive and precise applications of counter-messaging. Other studies show that many of the misinformed are likely to have already encountered and rejected correct information that was discomforting to their self-concept or worldview.24 Layered on top of this complexity are the network dynamics common to all social networks, as well as the unique network dynamics of individual social networks, and the network dynamics of the various user communities that are the targets of disinforma¬ tion or counter-messaging. Within this context, mis¬ information dynamics are different than those for information deficits. Discrediting the sources of dis¬ information and imposing reputational costs can be effective, but this is an immensely imposing challenge in an environment of computational propaganda and distributed, anonymous and easily replicable bot net¬ works. In short, this is a highly complex problem with asym¬ metric challenges. Computational propagandists do not require well researched articles or precise tar¬ geting of messages. They can spam disinformation through bot networks and see what works. Failures impose few costs. Meanwhile, the media, govern¬ ments and others who trade in truth require signifi¬ cant investments in researching and presenting that truth, and slight mistakes can generate mistrust and imperil reputations. What is clear is that the Department of State and public diplomacy professionals must carefully exam¬ ine and address the problem of computational pro¬ paganda from a multidisciplinary approach. This will require elements of cognitive psychology, network and influencer analysis, effective content creation, and the use of machine-driven communication tools and artificial intelligence systems. To accomplish this mission, the Department should commission a comprehensive assessment and evaluation of the literature of the science of persuasion, disinforma¬ tion and counter-messaging in the context of mod¬ ern information and communication technologies. It should assess technology tools available for com- ACPD I Understanding the Psychology Behind Computational Propaganda 23 batting computational propaganda and consider de¬ sired new tools. The assessment should identify key gaps in knowledge as well as promising areas for academic and practical experimentation. Finally, this assessment should drive a set of specific recommen¬ dations and best practices for public diplomacy pro¬ fessionals, both generally and situationally, that they can apply in their strategic planning and daily work tocountertheeffects of computational propaganda. ENDNOTES 1. MADCOMs can include many autonomous tools, including: robo-dialing telemarketing systems; web robots; Al chatbots like Zo and Xiao- ice; home assistants like Amazon Echo; and cutting edge (and mysterious) Als like "T" that dynamically create thousands of fake news videos for YouTube. 2. For more information on computational propaganda, see http://politicalbots.org/ 3. See Stephen G. Harkins and Richard E. Petty, "The Multiple Source Effect in Persuasion: The Effects of Distraction," Personality and Social Psychology Bulletin, Vol. 7, No. 4, December 1981; and The Multiple Source Effect: https://en.wikipedia.org/wiki/Social proof#Multi- ple source effect 4. See Harkins and Petty, "Information Utility and the Multiple Source Effect," Journal of Personality and Social Psychology, Vol. 52, No. 2, 1987; and Joseph W. Alba and Howard Marmorstein, "The Effects of Frequency Knowledge on Consumer Decision Making," Journal of Consumer Research, Vol. 14, No. 1, June 1987. 5. See Andrew J. Flanagin and Miriam J. Metzger, "Trusting Expert- Versus User-Generated Ratings Online: The Role of Information Volume, Valence, and Consumer Characteristics," Computers in Human Behavior, Vol. 29, No. 4, July 2013. 6. See Chanthika Pornpitakpan, "The Persuasiveness of Source Credibility: A Critical Review of Five Decades' Evidence," Journal of Applied So¬ cial Psychology, Vol. 34, No. 2, February 2004; and Source Credibility: https://en.wikipedia.org/wiki/Source credibility#Source credibility 7. See Cialdini, Robert (2009). Influence: Science and Practice. Boston, MA: Pearson Education, 114-166; and Implicit Egotism: https:// en.wikipedia.org/wiki/lmplicit egotism 8. See Dean, Jeremy (2007). "Why We All Stink as Intuitive Psychologists: The False Consensus Bias". PsyBlog; http://www.spring.org. uk/2007/ll/why-we-all-stink-as-intuitive.php; and False Consensus Effect https://en.wikipedia.org/wiki/False_consensus_effect 9. See Cialdini, R. B. (2016). Pre-Suasion: A Revolutionary Way to Influence and Persuade. New York: Simon & Schuster, 192-208. 10. This author believes that radicalizing young men online may be a goal of some state-sponsored trolling groups. These trolls use bots to create the appearance that hate speech is common, so that new online users find this behavior normal and mimic it. 11. See Michael G. Hughes, Jennifer A. Griffith, Thomas A. Zeni, Matthew L. Arsenault, Olivia D. Copper, Genevieve Johnson, Jay H. Hardy, Shane Connelly, and Michael D. Mumford, "Discrediting in a Message Board Forum: The Effects of Social Support and Attacks on Expertise and Trustworthiness," Journal of Computer-Mediated Communication, Vol. 19, No. 3, April 2014; Flanagin and Metzger, 2013; and Hughes et al., 2014. 12. For an excellent overview of the above techniques in the context of Russian disinformation, see Christopher Paul and Miriam Matthews, "The Firehose of Falsehood" December 2016, http://www.rand.org/content/dam/rand/pubs/perspectives/PE100/PE198/RAND_PE198. pdf 13. See Saul McLeod, "Moscovici and Minority Influence," 2007, https://www.simplvpsvchology.org/minoritv-influence.html; and Conversion Theory of Minority Influence: https://en.wikipedia.org/wiki/Conversion theory of minority influence 14. Cialdini, Robert (2009). Influence: Science and Practice. Boston, MA: Pearson Education, 208-236. 15. See Hasher, Lynn; Goldstein, David; Toppino, Thomas (1977). "Frequency and the conference of referential validity" (PDF). Journal of Verbal Learning and Verbal Behavior. 16 (1): 107-112; and The Illusory Truth Effect: https://en.wikipedia.org/wiki/lllusory truth effect 16. See Brendan Nyhan and Jason Reifler, "Misinformation and Fact-checking: Research Findings from Social Science," January 2012, New America Foundation. http://www.dartmouth.edu/~nyhan/Misinformation and Fact-checking.pdf ; and Belief Perseverance: https:// en.wikipedia.org/wiki/Belief perseverance; and Motivated Reasoning: https://en.wikipedia.org/wiki/Motivated reasoning 24 ACPD I Can Public Diplomacy Survive the Internet? 17. See Brendan Nyhan and Jason Reifler, "When Corrections Fail: The persistence of political misperceptions", 2010. Political Behavior 32(2): 303-330. https://www.dartmouth.edu/~nyhan/nyhan-reifler.pdf; and Joe Keohane, "How facts backfire", Boston.com, July 11, 2010, http://archive.boston.com/bostonglobe/ideas/articles/2010/07/ll/how facts backfire/?page=full 18. See, Brendan Nyhan and Jason Reifler, "The roles of information deficits and identity threat in the prevalence of misperceptions", February 24, 2017, https://www.dartmouth.edu/~nyhan/opening-political-mind.pdf 19. See Thorson, Emily, "Belief Echoes: The Persistent Effects of Corrected Misinformation" (2013). Publicly Accessible Penn Dissertations, http://repository.upenn.edu/edissertations/810/ 20. http://scienceblogs.com/cortex/2010/07/13/political-dissonance/ 21. See accompanying article by Matt Chessen describing how Al will transform machine driven communications 22. Thorson, Emily, "Belief Echoes: The Persistent Effects of Corrected Misinformation" (2013). Publicly Accessible Penn Dissertations. 810. http://repository.upenn.edu/edissertations/810 23. See Brendan Nyhan and Jason Reifler, "The roles of information deficits and identity threat in the prevalence of misperceptions", February 24, 2017, https://www.dartmouth.edu/~nvhan/opening-political-mind.pdf; Kuklinski, James H. and Paul J. Quirk (2000). "Reconsidering the rational public: cognition, heuristics, and mass opinion." In Arthur Lupia, Mathew D. McCubbins, and Samuel L. Popkin, eds., Elements of Reason: Understanding and Expanding the Limits of Political Rationality. London: Cambridge University Press, http://citeseerx.ist.psu. edu/viewdoc/download?doi=10.1.1.94.4875&rep=repl&type=pdf 24. See Cohen, G. L., Aronson, J., & Steele, C. M. (2000). When Beliefs Yield to Evidence: Reducing Biased Evaluation by Affirming the Self. Personality and Social Psychology Bulletin, 26, 1151-1164. ACPD I Understanding the Psychology Behind Computational Propaganda 25 i Jasor i Rais h - Illustrator RETHINKINGCOUNTERMEASURESINTHE AGE OF COMPUTATIONAL PROPAGANDA By Tim Hwang, Executive Director, Pacific Social From the rise of radio to the global adoption of social media, changes in technology have always powerful¬ ly influenced the landscape of communication. As a task that requires a nuanced and deft navigation of that landscape, public diplomacy is shaped by the available channels of mass communication, the dis¬ tribution of access to those channels and the favora- bility of those channels to certain styles of messaging defined by technology. Developing a strategy to effectively conduct public diplomacy and to counter emerging threats requires an understanding of the continuous technological shifts under way and the dynamics of communication that it gives rise to. This paper represents a preliminary attempt to artic¬ ulate how one emerging technological phenomena might impact the strategic doctrine of public diplo¬ macy in the modern era. Specifically, it takes up the phenomena of “computational propaganda” — the increasingly prominent combination of automation, sophisticated hoaxing and targeted messaging by state and non-state actors to manipulate discussion and spread misinformation online. By thinking holis¬ tically about these techniques and their objectives, it seeks to provide a framework for characterizing the nature of the threat they create, and the role of public diplomacy in responding to the challenge. Part I will examine the current and likely future of play, giving an account of how state and non-state ac¬ tors are leveraging computational propaganda. Part II will make an assessment of the threat, character¬ izing it as distinct in a number of ways from previous generations of strategic persuasion. Part III will turn to proposing a set of strategic principles defining the changing nature of communication and the conflicts of ideas taking place on online platforms. Part IV will then extend this doctrine to tactics, arguing that this analysis informs the potential approaches that might be used in countering these techniques. Finally, I will conclude by discussing further avenues for research and technological development. COMPUTATIONAL PROPAGANDA: A BRIEF OVERVIEW Recent years have seen the emergence and increas¬ ing sophistication of campaigns designed to manipu¬ late political discourse and suppress dissent. These efforts, referred to here as “computational propa¬ ganda,” have blended together a set of common components, merging automation with sophisticat¬ ed hoaxing, targeted messaging and cyberattacks in support of their objectives. Bots — fake user accounts that often autonomous¬ ly repeat the same or meaningless content — have figured prominently. Activists in Turkey and Syria have been subject to bot spamming campaigns that attempt to drown out oppositional political speech occurring on popular Twitter hashtags. In the United States, false accounts have been used to bolster the apparent grassroots support of political candidates, sometimes with a particular emphasis among key constituencies. In Mexico, one recent presidential election cycle featured two opposing groups of bots attempting to contest the other on social media. Bots were also a prominent feature of the online politi¬ cal discussion around the “Brexit” vote in the United Kingdom, helping to rally support around the deci¬ sion to leave the European Union. Fakeidentitiesareonlypartofthepicture. Beyond bot accounts, efforts havealso been uncovered thatinte- grate these methods as just one component of more sophisticated strategies for shaping public opinion. Oneexample detailed in the New York Times in 201 5 is the Russian “Internet Research Agency,” which has been connected with elaborate misinformation schemes that include fabricated videos and realistic clones of actual news sites. These types of actions are not limited to comparatively well-resourced gov- ernmentagencies. Bloomberg reported in April 2016 the story of Andres Sepulveda, who was involved in a series of private efforts to sway elections in Latin America through a combination of bots, compromise of voting machines and digital eavesdropping, among other techniques. 28 ACPD I Can Public Diplomacy Survive the Internet? Our existing knowledge of these campaigns of com¬ putational propaganda set a baseline in the sense that they indicate strong interest from state and non-state actors in engaging in these tactics. To the extent that we expect these campaigns to continue to improve and become more sophisticated, it is pos¬ sible to make projections along what dimensions we might see them develop in. Two technological trends seem particularly poised to bolster these tactics, namely, the recent breakthroughs in the field of arti¬ ficial intelligence (Al), and the development of better quantitative models of human group behavior. Machine learning (ML) — the subfield of Al research focused on the study of algorithms that improve themselves from data — has seen a rapid pace of de¬ velopment in recent years, driven by advancements in computing power and the availability of data. This has produced major breakthroughs in the capabil¬ ities of machines to accomplish many tasks previ¬ ously believed to be difficult to automate — from the ability to recognize objects in images and translate languages, to the operation of automobiles and mas¬ terful play of the game “Go.” In the context of compu¬ tational propaganda, these research developments may enhance the ability to create more believable fake identities and fabricated content in support of of these experiments remain the realm of academic inquiry, the results are openly published and might in¬ form the deployment of computational propaganda going forward. Future campaigns might more accu¬ rately model the behaviors of key influencers within a network to better shape their behavior, or better identify key moments for spreading information for maximal impact. Both trends point toward a space in which com¬ putational propaganda techniques become more effective at manipulating discussion online. These changes will increasingly create unique applications and threats which differentiate this phenomena from earlier methods of propaganda. SIMILAR OR DIFFERENT? From leafleting to radio broadcasts, the strategic use of persuasion by state and non-state actors to forward political objectives is nothing new. The tech¬ niques of computational propaganda are therefore not without precedent — they represent only the lat¬ est development in a historical legacy of methods meant to influence and manipulate discourse. To that end, many established principles in public diplomacy “...many established principles in public diplomacy can and will continue to apply in meeting the chal¬ lenge posed by this new breed of propaganda. ” campaigns of misinformation. Recent experiments have demonstrated the ability for ML to create realis¬ tic simulations of faces, including those of celebrities and world leaders. ML is also enabling the design of better conversational software, which can communi¬ cate believably with a human and potentially be lever¬ aged to increase the apparent credibility, authenticity and persuasiveness of a bot account online. Another advancing frontier of research is quantita¬ tive social science — an emerging field focusing on leveraging data available about social behavior to de¬ velop insights into how groups behave at scale. This field is generating interesting results, from develop¬ ing mathematical models for how ideas become pop¬ ular online, to understanding the design factors that might influence people to turn out to vote. While many can and will continue to apply in meeting the chal¬ lenge posed by this new breed of propaganda. However, a commonality of historical purpose across techniques should not distract from the ways in which new techniques might shape the costs, risks and op¬ portunities that actors face when choosing how to manipulate public opinion. In this respect, computa¬ tional propaganda appears to differ in three major ways from earlier methods of strategic persuasion. First, campaigns of computational propaganda can be conducted at significantly lower cost than per¬ suasion campaigns of the past. As easily replica¬ ble software, large numbers of bots can be quickly generated for a low investment and little technical infrastructure. Even groups with little technical ex- ACPD I Rethinking Countermeasures in the Age of Computational Propaganda 29 pertise may be able to quickly acquire bots and com¬ promised accounts through purchase online. Earlier strategies relying on printed media, or control over communications infrastructure, were comparatively more expensive. Second, the speed and geographic scope of per¬ suasion operations are considerably augmented by computational propaganda. Social media provides a channel whereby a persuasion campaign can im¬ mediately begin efforts to build trust and message to targeted groups throughout the world. While limited by internet penetration and the adoption of certain platforms, computational propaganda benefits from the global scope of social media. This provides a lev¬ el of access that would be cost-prohibitive to many organizations in an earlier communications environ¬ ment. Third, techniques of computational propaganda can be targeted and customized to a level of granularity greater than in the past. As opposed to a radio broad¬ cast or leafleting campaign, bots can customize their purported identity and their messaging to best take advantage of the biases and preferences of their tar¬ gets. This might take place on a level as granular as messaging tailored to an individual user online, using known data about that specific user to maximize the effectiveness of a persuasive effort, something diffi¬ cult to do effectively with earlier channels of commu¬ nication. Taken together, these are changes in degree that suggest a change in kind, particularly as computa¬ tional propaganda compounds shifts already under¬ way as a result of broader connectivity wrought by the spread of mobile devices and the global adoption of social media. These changes will inform the strate¬ gic doctrine which guides public diplomacy as it con¬ tinually evolves to meet a changing communications environment. AN EVOLVING STRATEGIC DOCTRINE The unique attributes of computational propaganda from earlier generations of strategic persuasion have several implications on the landscape of communica¬ tion. This informs an overall strategic doctrine — that is, a characterization of the nature of the challenge and the objectives of public diplomacy within it. For one, it appears that the generation of falsehood is poised in the near-term to enjoy ever increasing effectiveness at ever falling costs. Bots provide small groups the ability to influence at global scale at sub¬ stantially lower expenditure, and give well-resourced groups an affordable compliment to more elabo¬ rate influence efforts. At the same time, the costs of generating verified information and debunking false information remain relatively more expensive. This margin of cost between generating and countering falsehood may grow as the fabrication of realistic images, audio and video becomes cheaper and more accessible as a technology. In short, social media may produce the circumstances under which the of¬ fense of computational propaganda systematically has an edge against the defense of verification and fact-checking. Moreover, the emerging landscape is one in which it may be increasingly difficult to halt threats defini¬ tively and systematically. For one, commitments of public diplomacy to values of freedom of speech may exclude the use of approaches that attempt to drown out opposing discourse using similar “computation¬ al” methods. It is also difficult to avoid the constraints placed on governments by the fact that much of the focus of computational propaganda is on social me¬ dia platforms operated by private corporations. This may limit the ability for governments — particularly when the platform is based within the boundaries of another state — to obtain the needed data to effective¬ ly attribute and counter campaigns, and to constrain the open creation of new accounts through which to engage in computational propaganda. Multiplicity of the threat is also accompanied by the challenge of evaluating the importance of a given threat. Not only are computational propaganda cam¬ paigns able to leverage a wide range of attack vectors that make them difficult to detect, it is also difficult to assess whether a given campaign will have an im¬ pact on real events. Computational propaganda may occasionally serve as a distraction, creating exten¬ sive engagement and discussion on online channels without necessarily producing significant outcomes. However, what differentiates an ineffectual campaign from one that is a legitimate threat can be challeng¬ ing to ascertain at the outset. By the time success is evident, it may be too difficult to halt the momentum of the effort. In a context of limited resources, this as¬ pect of the landscape may present a major dilemma around how and when to deploy public diplomacy as¬ sets. These decisions are particularly difficult in the shadow of the so-called “Streisand effect,” in which efforts to directly confront a given message online it¬ self produces greater spread and discussion. 30 ACPD I Can Public Diplomacy Survive the Internet? The proliferation of persuasive power, augment¬ ed by trends in automation, machine learning and quantitative social science, will mean that a strate¬ gic posture purely aimed at counter-messaging to certain strategically important audiences will face greater challenges over time. A multitude of parties will control the means to engage in effective mass misinformation, with limited ability to detect, eval¬ uate and challenge each effort. Organizations with limited resources will quickly exhaust themselves confronting or containing falsehood, particularly as the believability of fabricated content continues to grow over time. Moreover, the available measures that have a possibility of categorically hindering these tactics are likely to conflict with other com¬ mitments towards preservation of the freedom of speech and media. As discussed above, the novelty of computational propaganda should not distract us from the reality ofitasjustthat — a form of propaganda. Tothatend, public diplomacy should continue to apply estab¬ lished techniques for confronting coordinated cam- paignsofmisinformationandstrategicpersuasion. Nevertheless, the new communications landscape and the computational propaganda techniques evolving within it introduce considerations that should join these established methods. Specifically, it may be critical to bring a “counter-networking” ap¬ proach to accompany existing efforts around count¬ er-messaging. In this context, the goal of public diplomacy would not be to defeat a specific narrative or propaganda campaign, but instead to ensure the robustness of the marketplace of information online. This would be both an offensive and defensive agenda that fo¬ cuses on the pattern of connections between social groups online. Defensively, this strategy would fo¬ cus on producing patterns of information exchange among groups that would make them difficult to effectively sway using techniques of computational propaganda. Offensively, the strategy would seek to distribute the costs of counter-messaging broadly, shaping the social ecosystem to enable alternative voices to effectively challenge campaigns of misin¬ formation. TAKING A COUNTER-NETWORKING APPROACH In order to be valuable, strategic doctrine should have a direct influence on tactics. How might a new focus on counter-networking manifest in concrete approaches to resisting the spreading use of com¬ putational propaganda? To make the approach more tangible, here are a set of different potential routes for exploration : Network topology: In a counter-networking context, it may be critical to adopt from the language of so¬ cial network science, which provides a collection of key metrics for summarizing the patterns of con¬ nections between individuals. Publicdiplomacy ini¬ tiatives might be targeted at shifting these metrics in a more discrete, tactical way, such as the average number of direct “friends” or “followers” possessed by an individual in the network, orthe average num¬ ber of links between any two individuals of the net¬ work. Raising or lowering these numbers strategi¬ cally within segments of users on a social network may give a more influential voice to allies and limit theoverall impact of misinformation efforts. Leveraging automation: Bots may be used in a myr¬ iad of ways beyond simply direct counter-messag¬ ing. It may be possible to leverage bots to signal to allies that misinformation is spreading, working to rally them to engage in a dialogue and push back online. Similarly, swarms of bots might be used to bridge connections between social groups not reg¬ ularly communicating online, helping to break echo chambers and diversify points of view. Doing so in a targeted way may create network topologies which “...the goal of public diplomacy would not be to defeat a specific narrative or propaganda cam¬ paign, but instead to ensure the robustness of the marketplace of information online. ” ACPD I Rethinking Countermeasures in the Age of Computational Propaganda 31 are more or less favorable to particular positions or feature a more ubiquitous distribution of individuals willing to actively challenge misinformation. Tools vs. messages: In the persuasive landscape formed by social media and computational propa¬ ganda, it may be attimes more effective to build tools, rather than construct a specific message. Apps that signal to users that a coordinated persuasive effort is taking place targeting them, for instance, may be more effective than an effort that attempts to chal¬ lenge each misinformation campaign as it is detect¬ ed. Similarly, open tools that make it easierto identify and debunk certain kinds of fabricated content may distribute the costs of identifying and investigating cases as they appear. These efforts are just a beginning. Counter-network¬ ing approaches may also significantly shape the collection of talents that are needed to conduct ef¬ fective public diplomacy. Adopting the three tactical suggestions discussed above would necessitate the development of more nimble software development, quantitative social science and machine learning ca¬ pability within organizations combatting computa¬ tional propaganda. CONCLUSION Ultimately, this essay represents a preliminary anal¬ ysis that is intended as an initial jumping off point for discussion. Whatever the eventual strategic approach, the rise of computational propaganda should not necessarily be seen as grouping of similar case studies or a blend of tactics being deployed by a specific adversary, but instead as a symptom of a changing landscape of communication. Such a frame enables thinking that looks beyond im¬ mediate challenges towards developing a general ap¬ proach to characterizing these threats and a ground¬ ed approach to addressing them. The falling cost and rising effectiveness of misinformation campaigns online multiply threats in a manner that makes a sin¬ gular focus on counter-messaging to specific audi¬ ences less efficient over time. To that end, traditional techniques might be bolstered by data-driven, count¬ er-networking efforts that seek to shape the patterns of connections between users online into topologies that are less susceptible to misinformation efforts and that help shoulder the costs of counter-messag¬ ing. This might manifest across a number of different countering tactics, each of which require further development and may involve greater collaboration with researchers not typically within the public diplo¬ macy community. However, as methods of compu¬ tational propaganda continue to proliferate and im¬ prove, such partnerships may become increasingly critical to meet the rapidly moving challenges posed by these techniques. PUBLIC DIPLOMACY’S (MISUNDERSTOOD) DIGITAL PLATFORM PROBLEM By Sam Ford, Research affiliate and consultant with Massachusetts Institute of Technology’s Pro¬ gram in Comparative Media Studies/Writing In 2013, an academic book I co-authored with Henry Jenkins and Joshua Green, Spreadable Media: Creating Value and Meaning in a Networked Culture, was published.1 The book examines the shifting business and cultural dynamics of a media environment where a signifi¬ cant portion of the public plays a more active, every¬ day role in the circulation of media texts. Specifically, we examined: “An emerging hybrid model of circulation, where a mix of top-down and bottom-up forces determine how material is shared across and among cultures in far more participatory (and messier) ways. The deci¬ sions that each of us makes about whether to pass along media texts — about whether to tweet the latest gaffe from a presiden¬ tial candidate, forward a Nieman Marcus cookie recipe email, or share video of a shoplifting seagull — are reshaping the me¬ dia landscape itself.” Spreadable Media lays out how changes in the media landscape had already been, and would be, affecting strategic communication professionals, emerging in¬ dependent media producers, the audiences for those texts, and our greater communication landscape as a result — particularly the international communication landscape. If anything, the past four years have shown just how rapidly the media landscape shifts. Marketing and public relations practitioners have poured budgets into tools to gather data on and — in some cases — qualitatively listen to discussions happening among their customers. Entertainment properties have moved rapidly to find new models to satisfy audienc¬ es looking to engage with content on-demand. And newsrooms have put significant emphasis on track¬ ing, and encouraging, sharing of their stories online. But, as the past four years have progressed, I have also watched developments unfold that we did not focus deeply in the book. For instance, the depth with which the algorithms of social network portals and dividuals, or the ways in which social bots can drive how content spreads (or does not spread, as the case may be). And I have consistently encountered new environ¬ ments and challenges beyond those we explored in Spreadable Media — such as the pressing question of how these shifts in the media landscape affect how effective public diplomacy is achieved across varied national and cultural contexts. In this piece, I reflect on how my research of, and ex¬ perience in, the media realm may have direct affect on those carrying out public diplomacy, with particu¬ lar focus on how emerging business models for digi¬ tal platforms threaten the integrity, not only of those platforms, but of the messages — including U.S. gov¬ ernment messages — therein. USING METAPHORS PAST THEIR DUE DATE In retrospect, perhaps my biggest regret from Spreadable Media is that, by demonstrating the impact that everyday people are having on what circulated, how it circulated, we inadvertently overly empha¬ sized examples where things spread fast, and far, when some of the most impactful, enduring content that spreads does so deeply, within and around par- ticularcommunities. This isto say, wemay haveover- emphasized pure virality without fully appreciating the importanceoflongevityand context. This reflects a problem that the media industries, industries with which governments and diplomatic actors are increasingly intertwined, are confronted with today. The business models of industries like television, radio and newspaper/magazine publishing were created at a time where companies had limited ability to understand the audiences they reached. As a result, these industries sought an advertising-sup- ftfW?werin9 two Simple search engines shape the circulation practices of in- 2 34 ACPD I Can Public Diplomacy Survive the Internet? Yet, as media professionals shift away from the mass media era of the 20th century, they have failed to shed outdated assumptions and models embedded within the industry. Instead, media companies have been hard at work trying to make current realities fit the industry architecture they are all so deeply in¬ vested in. Take, for instance, the power of the phrase “going viral” — a metaphorthat has been particularly attrac¬ tive to media organizations and marketers because it is defi ned by reaching that mass-scale audience that a previous media erawasabletoconvene. Even if the phrase, in most cases, does not accurately describe the cultural phenomenon it intends (people typically are making a range of active choices when choos¬ ing whether to watch/read/listen to something and then share it, as opposed to how we typically spread viruses amongst ourselves without intent or event awareness), the scale of “virality” has made it such a powerful metaphorthatevenaudiencesnowuses it. Virality brings with it the illusion that content can somehow be self-propagating. Television ratings, on¬ line traffic rankings, demographic segment profiles, focus group results and various other creations of the mass media world come to haunt us. Often these benchmarks of virality are applied with little acknowl¬ edgment that they were convenient shorthand — not reflections of reality — intended to make sense of a messy world that then outlived their usefulness. CLINGING TO OUR BUSINESS MODELS Despite the new possibilities offered by today’s digital platforms, the journalism and strategic communica¬ tion industries are still by and large governed by the “who are they, and how many of them are there?” business model. This approach remains driven by metrics of reach, clicks, shares and views, which ties success (either in terms of remuneration or strate¬ gic influence) to breadth and a logic of scale, particu¬ larly within target demographics. Governments, too, emphasize these types of metrics in demonstrating the reach of their strategic communications cam¬ paigns and outreach initiatives. For instance, Russia’s RT touts itself as the most viewed news channel on YouTube, despite the fact that the content driving its popularity, if authentic, is largely apolitical. Similarly, the Voice of America heavily emphasizes its weekly reach as proof of its continued relevance, even if oth¬ er markers may be far better proof of its impact and value. In short, as we have entered a new communication landscape with powerful possibilities to tell new sto¬ ries in new ways, and to support them in a global market, the primary media platforms we depend on have preserved the business logic of the broadcast world. And the rise of big data makes this focus on what is most easily counted more central than ever.3 Much of this continued support of old advertising models has been exacerbated by “investor story¬ time.”4 Ethan Zuckerman explains, “investor story¬ time is persuading investors that your ads will be worth more than everyone else’s ads. That is because most online ads are not worth very much.”5 This means each of the primary digital platforms — Face- book, Google, Twitter and the like — target the same programmatic ad budgets. Convincing investors to support a new digital platform or publication is ul¬ timately about explaining why your “viral engine” is going to lead to better traffic, or why your native sur¬ veillance techniques allow improved target advertis¬ ing, or about how you can deliver an audience more cheaply than competitors. As Joe Marchese — currently president of advanced advertising for Fox Networks Group — writes that one of the fundamental problems is that new “impres¬ sions” are created in a digital landscape, without nec¬ essarily any real gain in human attention.6 In short, platforms can add auto-play videos to people’s social feeds, a new advertising spot on a page, or an extra ad to pre-roll and count it as new “impressions,” even if there is no actual gain in meaningful attention from an audience. It is as if people forgot that social media “It is as if people forgot that social media met¬ rics are proxies for something else, and instead started taking their own creations literally. ” ACPD I Public Diplomacy’s (Misunderstood) Digital Platform Problem 35 metrics are proxies for something else, and instead started taking their own creations literally. It is one thing to employ storytime for the people who you are trying to get to pay you. It is another to use it on your¬ self. NOT ENOUGH DISTANCE This tail-chasing of outdated metrics is not merely a waste of financial and human resources. By relying on platforms grounded in faulty business logic, orga¬ nizations, governments and companies risk severely compromising their brands and credibility too. One problem is that well-known, fact-based news¬ rooms and storytellers rely on the exact same media hubs (e.g, Facebook, Twitter, etc.) as thin news and openly partisan sites to distribute their content. But these hubs are primarily focused on click-throughs, or when a user clicks on an advertisement, which can then be leveraged into additional or more profitable ad sales. Publishers, both benign and malicious, are thus forced to compete with one another, not based on the quality of their content, or the longevity of their brands, but by using catchy and sensational headlines to try and lure views from users as they dash through the social media feed. These are, incidentally, the ex¬ act same tactics utilized by tabloid publications like the National Enquirer to grab attention at a check-out counter, only scaled to every smart phone and com¬ puter in the world. And now journalists and scam artists are competing with one another for your at¬ tention, side-by-side. The only difference is that sites disseminating thin news or outright disinformation do not have to pay the overhead that comes along with producing fact-based and deeply contextualized news. Consider, for instance, one tactic dubbed by Variety’s Andrew Wallenstein as a “controversial practice” whereby digital publishers are “inflating traffic num¬ bers”7 by counting traffic that does not come from its own sites. Rather, publishers strike deals with in¬ dependent sites to sell their ad inventory alongside the ad space on theirown sites. So, when a publisher talks about its audience reach, it is actually including the traffic of sites it does not own or control . Or take into account the ways in which journalism sites engage in business practices that deeply inter¬ twine them in a tangled web of digital publishing with non-news current event sites, openly partisan sites, scandalous articles and blatantly misleading articles. Often, through “content amplification windows” that provide paid links to stories on other sites, deeply researched news articles conclude with links to ar¬ ticles from publications with much lower editorial standards, or even disinformation sites. And, in reverse, look at the various ways digital pub¬ lishers engage in paid promotion to drive traffic to their stories, for instance, by paying to get a link to a legitimate news story on a non-news or even an in¬ tentionally misleading site. But why? According to Lucia Moses, “publishers need to show big numbers,” which “rewards tricks to inflate the size of their audiences and to make them appear younger than they actually are.”8 This means we have created a dynamic where there is not as much distance as there should be between the business practices of legitimate journalistic or¬ ganizations and the realm of scam artists and spam¬ mers. As Sean Blanda has written, “The methods used to fund modern journalism simultaneously un¬ dermine trust in the news outlets... News publications aren’t (or can’t afford to be) policing their ads. Seedy brands are literally stealing the credibility of news sites for a few pennies.”9 One of the fundamental reasons this has happened is that media industries and advertising metrics do not seem to account for negative ROI, or the erosion of trust that may result from clickbait headlines and rankings that highlight gross/unique views while ig¬ noring bounce and completion rates.10 The scam art¬ ist is not concerned about “negative ROI,” because the goal is not to build a permanent, lasting brand or trust with citizens. This dynamic is further driven by the fact that legiti¬ mate journalism enterprises in an online setting are often selling their ad inventory based on “Who are they, and how many of them are there?” In an envi¬ ronment where impressions are commodified, all the infrastructure of running a legitimate news oper¬ ation starts to appear as a drag on the bottom line, with no upside. If you do not focus on a business mod¬ el wherein having a dedicated readership, a trusted relationship with an audience, or a publishing brand that means something over time, then organizations whose standards for publishing something is lower, or who make no pretense for building a long-term following and brand, might be able to generate the semblance of enough traffic to compete on volume by engaging in many of the same tricks, with much lessoverhead. 36 ACPD I Can Public Diplomacy Survive the Internet? Journalism brands, driven by a concern about monthly traffic patterns, at the exclusion of building a long-term brand, have business models that give no value to their strongest asset. As my former Fusion colleague wrote in 201 6, “If you have lots of traffic but little brand value, then you can disappear more or less overnight: look at Upworthy. On the other hand, if you have low ratings but a strong global brand, then you can still be worth a fortune: look at CNN.”11 Yet, it is a common concern for traffic that leads news or¬ ganizations to continuously make choices that sacri¬ fice long-term trust for the daily, weekly, or monthly traffic goals — leading to clickbait-style headlines that sometimes do not sound as different as they should from, say, The Washington Post, The Huffington Post, and The Gummy Post (examples only chosen because of what different “posts” each should have). Writes Blanda about misleading headlines from legitimate news sources, “The problem isn’t that news outlets make these mistakes. It’s that they make them be¬ cause they have business incentives to do so.”12 The current model not only does a disservice to jour¬ nalism and the public, but it often leads news organi¬ zations awry from their missions and seeks to create deceptive metrics, artificially inflating the broadest audience possible for advertisers. If the current model contributes to the erosion of trust across our shared digital platforms, then this creates a funda¬ mental problem for anyone relying on the digital spaces to connect with audiences, including the pub¬ lic diplomacy community. CONCLUSION: THE RAMIFICATIONS FOR PUBLIC DIPLOMACY Diplomats rely on trusted, rigorous journalism insti¬ tutions to provide more credible verification of facts. However, the damaging business practices erode the authority independent corroboration from a news organization can provide. And they make it even more prone for various actors — from other governments, private industries and overtly politically partisan players — to question the veracity of information writ- large. As rigorous journalism and quasi-news sourc¬ es commonly adhere to the same business models and engage in many of the same business practices, we run the risk of driving audiences from healthy skepticism into cynicism and moral relativism.13 Yet, diplomats should be concerned about more than just how these tactics are undermining the credibility of legitimate news brands. As the U.S. government pays $1 .8 billion a year in taxpayer dollars to deliver our messages to audiences around the globe, public diplomacy practitioners are increasingly dependent on the same media ecosystem described throughout this essay. While news organizations are controlled by an advertising-driven business model that needs to bring revenue in, public diplomacy is not. Yet, like strategic communication teams in various sectors, it is easy to fall into the same trap of measuring success heavily by reach, clicks, shares and views, and shap¬ ing initiatives, content and traffic around what best reaches those goals. For-profit media companies will likely continue to struggle to find ways to appropriately value engage¬ ment depth, completion rates, story shelf life and their long-term brand strength, among other factors, over the breadth that dominates programmatic dig¬ ital ad buying. However, there is no reason that the same problems must plague how public diplomacy is measured and evaluated. As many of the commercial forces behind digital publishing and sharing continue to shape an environment that lead publishers down a path of diminished reader trust, public diplomacy practitioners should be careful not to head down the same trail. Instead, the should develop publishing and circulation strategies that are measured by meaning¬ ful metrics and be careful to avoid the traps in which for-profit publishers and platforms currently find themselves. By taking into account the potential for negative ROI that comes along with many methods of achieving reach and scale, and by finding ways to measure success through where and how messag¬ es are meaningfully resonating with their audienc¬ es, the State Department will be better prepared to safeguard its content and brand from tactics meant to challenge the credibility of their messages and en¬ gagements around the world. ACPD I Public Diplomacy’s (Misunderstood) Digital Platform Problem 37 ENDNOTES 1. Henry Jenkins, Sam Ford, and Joshua Green, Spreadable Media: Creating Value and Meaning in a Networked Culture. New York: New York UniversityPress,2013. 2. Ilya Vedrashko, “How Spreadability Changes How We Think about Advertising,” Spreadable Media, 2013, http://spreadablemedia. org/essays/vedrashko/. 3. As we have more ways to gather data on audience, the illusion of certainty these numbers bring with them have made quantified reach even more central in the narratives of reach organizations tell the outside world and themselves. 4. Maciej Ceglowski, “The Internet with a Human Face,” talk at Beyond Tellerrand in Dusseldorf, Germany, on 20 May 2014, http:// idlewords.com/talks/internet_with_a_human_face.htm. 5. Ethan Zuckerman, “The Internet’s Original Sin, The Atlantic, 14 August 2014, https://www.theatlantic.com/technology/ar- chive/2014/08/advertising-is-the-internets-original-sin/376041/. 6. Joe Marchese, “Outside Voices: Beware The End of Advertising?” Wall Street Journal, 07 November 2014, https://blogs.wsj.com/ cmo/2014/11/07/outside-voices-beware-the-end-of-advertising/. The challenge is that adding new elements to your page — two new slots for ads on a page, for instance — may count as additional “impressions,” but the amount of attention the audience has to give has not increased, as it is a finite resource. 7. Andrew Wallenstein, “Vice Media Traffic Plummets, Underscoring Risky Web Strategy,” Variety, 21 March 2016, http://variety. com/201 6/digital/news/vice-media-traffic-plummets-underscoring-risky-web-strategy-1 201 733673/. 8. Lucia Moses, “The Favorite Publisher Trick that Makes Their Digital Audiences Look Bigger and Younger,” Digiday, 18 February 2016, https://digiday.com/media/favorite-publisher-trick-makes-digital-audiences-look-bigger-younger/. 9. Sean Blanda, “Medium, and the Reason You Can’t Stand the News Anymore,” Medium, 15 January 2017, https://medium.eom/@ SeanBlanda/medium-and-the-reason-you-cant-stand-the-news-anymore-c98068fec3f8. 10. The phrase “clickbait” refers to articles with headlines that are focused on trying to generate as much curiosity and clicks as pos¬ sible, even if they may be particular scandal-driven, partisan, or even misleading. Gross page views refer to the number of overall views a site’s pages have received in a particular time period, while unique views refer to the total number of people who have passed through the site in a given time. Bounce rate refers to the percentage of viewers who come to the site, only to instantly leave without spending enough time on the site to engage with any content. Completion rate refers to the percentage of a piece of content that the audience gets through before leaving. 11. Felix Salmon, “Why Messaging Is the Future of the News Brand,” Fusion, 1 1 February 2016, http://fusion.net/why-messaging-is-the- future-of-the-news-brand-1 793854684. 12. Blanda. 13. For more on my call to action to the news industry, see Sam Ford, “The Year We Talk about Our Awful Metrics,” Nieman Lab, Decem¬ ber 201 6, http://www.niemanlab.org/201 6/1 2/the-year-we-talk-about-our-awful-metrics/. Five years from now you won't have any idea whether you are interacting with a human online or not. In the future, most online speech, digital engagement and content will be machines talking to machines. UNDERSTANDING THE CHALLENGES OF ARTIFICIAL INTELLIGENCE ANDCOMPUTATIONAL PROPAGANDA TO PUBLIC DIPLOMACY By Matt Chessen, Foreign Service Science, Technology and Foreign Policy Fellow at The George Washington University Machine-driven communications tools are a reality now and the addition of emerging artificial intelli¬ gence (Al) tools will enable machines to dominate the online information space. This paradigm shift isn’t limited to artificial personal assistants like Siri and recreational chatbots like Xiaoice.1 It refers to ma¬ chine-driven communication overwhelming Face- book, Twitter, YouTube, Tinder, Snapchat, Reddit, chat rooms, news site comment sections and the rest of the social web. All of it will be dominated by machines talking. This machine communication will become nearly indistinguishable from human communica¬ tion. The machines will be trying to persuade, sell, deceive, intimidate, manipulate and cajole users into whatever response they are programmed to elicit. They will be unbelievably effective. There is an urgent need to think strategically about what this transformed information ecosystem means for the practice of public diplomacy (PD). In addition to discussions of competing with compu¬ tational propaganda efforts, PD practitioners need to consider the question of how they can create and sustain meaningful conversations and engagements with audiences if the mediums typically relied upon are becoming less trusted, compromised and domi¬ nated by intelligent machines. Put simply, we current¬ ly take for granted that there are trusted platforms in virtually every country we operate to directly engage with local audiences. The integrity of these platforms will be threatened, and we need to start thinking now about how to best engage foreign audiences in a world dominated, if not overwhelmed by, artificially intelligent computational propaganda. Also, utilizing new Al tools for public diplomacy will require a reinvention and reimagination of business processes that takes into account the speed, person¬ alization, autonomy and learning capabilities of Al systems. We cannot just replicate the same process¬ es using new tools. That would be a failure. Public di¬ plomacy must be completely reinvented for the 21st Century. MACHINES TALKING TO HUMANS TALKING TO MACHINES TALKING TO MACHINES Advances in artificial intelligence will soon enable highly persuasive machine-generated communica¬ tions. Imagine an automated system that uses the mass of online data and easily available marketing databases to infer your personality, political prefer¬ ences, religious affiliation, demographic data and interests. It knows which news websites and social media platforms you frequent and it controls multi¬ ple user accounts on those platforms. The system dynamically creates content — everything from com¬ ments to full articles — specifically designed to plug into your particular psychological frame and achieve a particular outcome. This content could be a collec¬ tion of facts, fake news or a mix of just enough truth and falsehood to achieve the desired effect. The Al system has a chatbot that can converse with you, through text, voice or even video. The chatbot will be nearly indistinguishable from a human being and will be able to operate in multiple languages. The Al chatbot will engage you in online discussions, debate you and present compelling evidence to persuade you. It could also use information from databases or social media to discover your weaknesses and use this information to troll you and threaten your family. The Al system will be able to detect human emotions as well or better than people can. Similarly, it will mimic convincing human emotions that resonate with your own personality and emotional state. It will be a learning machine, so it will figure out approach¬ es and messages that influence you the best. It will select for success and improve constantly. It will run A-B tests with people who share yourcharacteristics to determine what messages are most effective and then deploy those messages to similar populations. Like other digital tools, once created, the marginal cost of creating more is almost zero. So there could be millions of Al chatbots prowling the internet, 24 hours a day, seven days a week, vying for your atten- 40 ACPD I Can Public Diplomacy Survive the Internet? tion so they can infect your brain with their message and change your behavior. Systems looking for humans to influence will inevita¬ bly wind up trying to persuade other machine-driven accounts posing as humans. The machines will talk to, at and over each other, drowning out human con¬ versations online with a tidal wave of machine-driv¬ en speech and content. The online information envi¬ ronment will be overwhelmed with machine-driven speech designed to sell, persuade, intimidate, dis¬ tract, entertain, advocate, inform, misinform and ma¬ nipulate you. This is a highly probable vision for the information environment we will move into over the next sever¬ al years. Our actions now will shape whether spaces are preserved for democratic speech and discourse, or whether the social web will be destroyed by an in¬ vasion of highly intelligent machine driven communi¬ cation tools. Our uptake of these tools and redesign of PD business processes around new technologies now is the only way to ensure U.S. public diplomacy remains relevant in the future. MADCOMS AND COMPUTATIONAL PROPAGANDA The basic technologies for this evolving information environment are MAchine Driven COMmunication (MADCOM) tools. Theweb robot, or“bot,” is the most common typeof MADCOM. 2 Bot capabilities are lim¬ ited to providing basic answers to simple questions, publishing content on a schedule or disseminating content in response to triggers. However, bots can have a disproportionate impact because it is easy to create a lot of them and bots can post a high volume content at a high frequency. An individual can easily operate hundreds of Twitter bots with little techni¬ cal knowledge using readily available hardware and software. Bots are currently used by nations, corpo¬ rations, politicians, hackers, individuals, state-spon- soredgroups,NGOsandterroristorganizationsinan effort to influence conversations online. Computational propaganda is a new term for the use of machine-driven communication tools for polit¬ ical purposes. These purposes can range from rela¬ tively benign amplification of political messages to in¬ sidious state-sponsored trolling and disinformation. Currently, primarily simple (i.e., non-AI) bots are used for computational propaganda. These follower, road¬ block and propaganda bots are used for amplifying people and ideas, suppressing or diverting online speech, and more traditional influence operations. Emerging Al tools will radically enhance the efficacy of MADCOMs and computational propaganda tech¬ niques. (Note: For more information on computa¬ tional propaganda, see the accompanying papers by Matt Chessen, Samuel Woolley and Tim Hwang). ARTIFICIAL INTELLIGENCE WILL RAD¬ ICALLY ENHANCE COMPUTATIONAL PROPAGANDA EFFICACY Artificial intelligence (Al) popularly refers to an evolving constellation of technologies that enable computers to simulate cognitive processes, such as elements of human thinking. Al is also a discipline (like biology or chemistry) that is concerned with cre¬ ating machines that can make decisions well under uncertainty, perceive data or the environment, and act to satisfy some objective. Today’s Al is confined to specific tasks (“narrow” Al), like providing driving directions or recognizing faces in images, and is not a general intelligence applicable across many domains ora super-intelligence exceeding human abilities. Machine learning is a subset of Al. Machine learning extracts patterns from unlabeled data (unsupervised learning) or efficiently categorizes data according to pre-existing definitions embodied in a labeled data set (supervised learning). Machine learning is used in Google’s search algorithm, digital advertising and on¬ line personalization tools (e.g. the Amazon and Netflix recommendation engines; the Facebook newsfeed). Machine learning also extends into quantitative pro¬ cesses — such as supply chain operations, financial analysis, product pricing and procurement bid pre¬ dictions. Nearly every industry is exploiting machine learning applications. Deep learning isatypeof machine learningthatuses additional, hierarchical layers of processing (loose¬ ly analogous to neuron structures in the brain) and large data sets to model high-level abstractions and recognize patterns in extremely complex data. Deep learning systems manage very large data sets better than other Al tools and are ideal for understanding data-rich and highly complex environments.3 These tools are not confined to wealthy corporations or state-sponsored actors. Al tools are widely avail¬ able (Google’s T ensorFlow, M icrosoft’s Control T oolkit ACPD I Understanding the Challenges of A! and Computational Propaganda to PD 41 and many other Al tools are free and open-source) and operate on common computer hardware. HOW Al WILL TRANSFORM MACHINE DRIVEN COMMUNICATIONS Al chatbots are increasingly capable of engaging in robust conversations about complex topics. For example, Microsoft’s Mandarin language Al chatbot ‘Xiaoice’ has sophistication, empathy and conversa¬ tional flexibility that make “her” extremely popular. Xiaoice has 20 million registered users, average us¬ ers interact with her 60 times a month and she was ranked as Weibo’s top influencer in 2015. She aver¬ ages 23 exchanges per user interaction. That is not trivial experimentation; it is a conversation. Some us¬ ers relate intimately to Xiaoice and consider her an always-available friend and confidant.4 Currently Xiaoice requires a team of engineers to achieve this level of sophistication. This level of chat¬ bot technology is well within the capabilities of a corporation or nation-state, but still unavailable to the masses. However, like all digital technology, it will improve in capability and accessibility. Over the next several years, high-end chatbots like Xiaoice will become indistinguishable from humans in a broad tion. Emerging debating technologies7 will allow Al chatbots to persuasively argue by analyzing a corpus of knowledge, determining pro and con arguments, and creating dynamic, persuasive content in support of a position. Al tools are increasingly sophisticated at affective computing,8 one aspect of which is determining hu¬ man emotional states9 from text, facial expressions and vocal patterns. This will allow machines to inter¬ pret whether you are happy, sad, anxious, relaxed or open to a communication when they interact with you. Al tools can then tailor their communication to your mood with just the right amount of emotional emphasis to achieve the desired effect. If an affective Al tool detects that the target is impatient and doesn’t feel like conversing at the moment, the Al can cease communication and try messaging them later when they are more persuadable. If a target is curious and wants to talk politics, the Al will detect openness in their communications and can engage them in a live¬ ly conversation (or argument). If the Al detects emo¬ tional vulnerability, it could prey on those emotions to persuade, manipulate, or intimidate. In another twist on affective computing, scientists are training Als to accurately emulate human emo¬ tions10 in the facial expressions of avatars. This will “0 ver the next several years, high-end chatbots like Xiaoice will become indistinguishable from humans in a broad range ofconversations. ” range of conversations. When the technology prolif¬ erates, chatbots will converse fluidly with humans on platforms ranging from social media apps to news discussion boards to dating sites, about a wide vari¬ ety of topics. Altoolsarealsoimprovingatdynamically generating unique content and will soon be developing custom propaganda, disinformation and persuasive argu¬ ments. Currently, humans develop content for com¬ putational propaganda that is then distributed by bots. Al tools are already capable of generating be¬ spoke content, like news articles5 and novels,6 using predefined parameters. The quality of this content will improve and Al systems will be able to commu¬ nicate across more subjects with greater soph istica- be useful for generating custom, persuasive video, but the technology can also be used to alter reality and generate disinformation. Researchers at Stan¬ ford University have developed real-time facial re¬ enactment tools11 that allow users to take existing videos — like a speech by a world leader — and realis¬ tically manipulate the speaker’s facial expressions. The resulting videos12 show realistic, if not yet per¬ fect, manipulations of the speaker’s face and mouth. Concatenative speech synthesis,13 or better yet, voice conversion14 technologies like Google Deep Mind15 will allow machines to replicate anyone’s voice from samples. If combined with affective computing, fa¬ cial re-enactment tools and an Al chatbot, this would give propagandists the capability to create videos of anyone saying anything, or more insidiously, to subtly 42 ACPD I Can Public Diplomacy Survive the Internet? modify existing video for propaganda or disinforma¬ tion purposes. Affective computing allows the emo¬ tional inflection of an altered human speaker or a dy¬ namic Al MADCOM to be precisely tailored to achieve the desired influential outcome. Big data combined with machine learning tools will enhance the ability of MADCOMs to influence people through highly personalized propaganda. In the Unit¬ ed States alone there are several thousand data bro¬ kers. One company, Acxiom, claims to have16 an aver¬ age of 1 ,500 pieces of information on over 200 million Americans. Another company, Cambridge Analytica, claims to have 3,000-5,000 data points per individ¬ ual17 and psychological profiles18 on 230 million U.S. adults. We give away our data when we shop using supermarket club cards, when we browse the inter¬ net, when we take “fun” Facebook personality tests,19 and through hundreds of other seemingly innocuous activities. The spread of “Internet of Things” devices means a proliferation in the amount of data that could know us personally and intimately. The communica¬ tions generated by Al MADCOMs won’t be mass me¬ dia, they will be custom tailored to speak to an indi¬ vidual’s political frame, worldview and psychological needs and vulnerabilities. Because Als are learning systems, they improve rapidly with experience. An Al could autonomously determine which of its thousands of pieces of pro¬ paganda, disinformation or intimidation are most effective and emphasize or evolve those, while quick¬ ly ending failing campaigns. Al tools will test target weak points and learn what provokes the desired emotional response. By probing with multiple ac¬ counts and messages, an Al could learn that person¬ al threats to a particular journalist provoke little re¬ sponse, but threats to their loved ones provoke fear. So, the MADCOM Al could pose as members of a local hate group who threaten the journalist’s children un¬ til they stop reporting. And while that journalist might not be troubled by abusefromafew MADCOM trolls, “Since machines are not limited by human tem¬ poral constraints, they can operate 24/7/365 andrespondtoeventsalmostimmediately.” be captured about our lives. Virtual reality will give others the opportunity to test our actual reactions to hypothetical stimuli and to measure our responses to products and ideas subtly introduced into the back¬ ground of virtual experiences. Data breaches from private companies and government databases have exposed extremely private information about us and our associates. And we increasingly volunteer our most intimate details online, posting photos of family vacations and tweeting our opinions. Al tools could use all of this information to tailor persuasive, dis¬ tracting or intimidating speech towards individuals based on their unique personality and background. Human cognition is a complex system, and machine learning tools are very good at decoding complex systems. When provided rich databases of informa¬ tion about us, machines will know our personalities, wants, needs, annoyances and fears better than we know them ourselves. Machines will know how to in¬ fluence people who share our traits, but they will also an onslaught of threats from thousands of Al-driven accounts, most of which look and speak like people in their community, would significantly escalate the effectiveness of the campaign. Digital tools have tremendous advantages over hu¬ mans. Once an organization creates and configures a sophisticated Al chatbot, the marginal cost of run¬ ning that tool on thousands or millions of user ac¬ counts is relatively low. Since machines are not limit¬ ed by human temporal constraints, they can operate 24/7/365 and respond to events almost immediately. Once an Al is trained to understand a subject domain, it can be programmed to react to certain events with speech and content produced at machine speed, shaping the narrative almost immediately. Al tools will know key influencers and populations with per¬ sonality profiles or political inclinations that are susceptible to their messages. The Al systems will target additional vulnerable users with dynamical¬ ly generated communications instantly and in real ACPD I Understanding the Challenges of A! and Computational Propaganda to PD 43 time as events unfold. This is critical in an informa¬ tion environment where the news cycle is continually squeezed into smaller and smaller windows. Often, the first story to circulate is the only one that people recall, even if it is untrue. Research demonstrates20 that once a fake news story is believed, it is very diffi¬ cult to change people’s minds, even when presented with compelling contrary evidence. How can journalists, diplomats, public relations staff, politicians and government officials plan to compete with Al MADCOMs that can interpret and react to stories almost instantly, developing and deploying customized communications personalized to indi¬ viduals and groups before humans can even begin a first draft? How can a government press release, or a carefully crafted, researched and fact-checked news article, or a corporate public relations campaign, precisely developed over months, ever compete with real time, personalized, always available, dynamically generated, instantaneous, machine-driven manipu¬ lative speech, text, video and other content? The answer is: humans cannot compete alone. On digital networks, only humans teamed with Al ma¬ chines can compete with Al machines. The rise of Al-driven MADCOMs will spur an information arms race as empowered individuals, NGOs, corpora¬ tions and governments all strive to shape narratives around events. The “bad guys” will have their MAD- COM Als, and the “good guys” will have their own. Everyone will have Al tools that try to identify adver¬ sary MADCOM accounts. These attribution tools will be used to anticipate computational propaganda campaigns, respond to ongoing operations and dif¬ ferentiate human users from machine users. Similar to the cybersecurity struggle, the internet will be the battleground for a continual cycle of one-upmanship as technologists improve Al detection tools and pro¬ pagandists improve Al MADCOMs to avoid detection. The most sophisticated machine accounts will be nearly indistinguishable from the human accounts. But many propagandists may not bother with detec¬ tion tools since there is little marginal cost to spam¬ ming machines and people with speech and content. So, in a bizarre twist, machines will frequently run their information campaigns against other machines.21 Those targeted, machine-driven accounts will re¬ spond with their own communications and the online information space be swamped with machines ar¬ guing with machines. MADCOMs will overwhelm hu¬ man-generated speech and communication online. This raises a number of larger policy questions for governments and the private companies that make up the social communications infrastructure. Social media companies could adjust their tools and pol¬ icies to make it much harder for machines to oper¬ ate on their platforms.22 However, there are often financial incentives against restricting MADCOM ac¬ counts, depending on the platform. Prohibiting ma¬ chine-driven accounts could significantly change the nature of the business model and communication on the platform. Similarly, MADCOMs raise complicated questions where they are used by adversary govern¬ ments. These range from basic cybersecurity con¬ cerns to questions about the role of free speech and definitions of acceptable norms for state behavior in cyberspace. For countries that respect universal hu¬ man rights and democratic values, MADCOMs pose a threat to the integrity of discussion spaces used by citizens. But efforts to regulate them would certain¬ ly create unpredictable follow-on effects and should not be attempted without careful consideration. FROM COMPUTATIONAL PROPAGANDA TO COMPUTATIONAL DIPLOMACY: REC¬ OMMENDATIONS FOR PUBLIC DIPLOMACY PROFESSIONALS Awareness of MADCOMs, computational propagan¬ da and emerging artificial intelligence technologies is crucial for understanding the modern information environment. Information on these tools and tech¬ niques should be included in every level of training for PD professionals. This will allow practitioners to understand the dynamics of online communica¬ tions, identify where computational propaganda techniques are in play and effectively counter them. Practitioners should maintain awareness of new re¬ search in computational propaganda, Al, bots and related technologies.23 The Department of State should develop a compre¬ hensive strategy and plan of action for managing the impact of MADCOMs, analyze the possibility of utilizing MADCOMs in public diplomacy and intro¬ duce in-house Al tools where appropriate.24 The 44 ACPD I Can Public Diplomacy Survive the Internet? Department should consider the worst-case scenar¬ io and be prepared for the possibility that the online information environment may be completely overrun with machine-driven speech. This would have a sig¬ nificant impact across a broad range of diplomatic efforts that use the internet for messaging and com¬ munications. In a world dominated by machine-generated, self¬ learning and propagating content, maintaining the integrity of our communications campaigns and our commitment to truthful content with open attri¬ bution is crucial to the efficacy of our efforts. Within these parameters, the Department should develop its own in-house MADCOM and Al tools. This does not mean the department should engage in disinfor¬ mation efforts or unattributed propaganda. Like all technology, artificial intelligence can be used for both good and bad purposes. A number of tools and methods are available to help human public diplomacy practitioners counter pro¬ pagandists and their machines. The Department must continue to develop sophisticated Al detec¬ tion and attribution tools to identify and counter disinformation campaigns before they spread. The psychology behind computational propagan¬ da effectiveness indicates that counter-messaging established messages is unlikely to be effective.25 This implies that the best, and perhaps only way, to counter computational propaganda is to detect dis¬ information campaigns when initiated and develop rapid-response messaging campaigns to stifle the disinformation before it goes viral. Machine learning tools will be critically important in identifying and monitoring adversary bot networks. There may also be utility in “outing” bot networks to social media companies, but this may have the un¬ intended consequences of forcing propagandists to improve their tools or driving them to more diverse channels. As adversary MADCOMs become more sophisticated and integrate Al technologies, Al tools may be required for users to determine whether they are interacting with a human or a machine online. The Department should work with private companies and academia to make these tools available. The Department should develop Al chatbots. These chatbots could discuss U.S. foreign policy generally, or specific topics like countering violent extremism or assisting U.S. businesses abroad. These chatbots could be language localized and placed on U.S. em¬ bassy websites worldwide, or located on popular messaging platforms like Facebook Messenger or Kik. They can be an effective way to engage technolo¬ gy-friendly youth and provide a means for communi¬ cating policy ideas across a wide variety of platforms without direct human intervention. Debating systems will soon allow chatbots to have robust discussions about complex topics, including U.S. foreign policy. The Department should develop Al scanning tools that improve on keyword searches by autonomously scanning for conversations and content relevant to U.S. foreign policy and flagging those items for action. When relevant conversations are identified, humans could intervene with a conversation or content, or Al chatbots could be tasked with autonomously engag¬ ing users in discussions that promote U.S. policies or ideas. Short of interventions, using Al machines to simply track these conversations on the internet can be helpful in providing real-time insights into opin¬ ions about U.S. foreign policy, unearthing nuances between key audiences in certain countries, and en¬ abling careful thinking about how to properly articu¬ late American policies to these same audiences. The Department should investigate tools for auton¬ omous content creation regarding U.S. policy posi¬ tions. Al’s are already capable of writing content for sports and earnings reports that is indistinguishable from human generated articles. This capability will increase significantly over the next several years. Al tools will soon be able to write first-drafts of speech¬ es, create press releases and generate text, images and video for social media faster than a human can begin to consider a first draft. Such tools may be¬ come crucial as the Department is asked to accom¬ plish more with fewer resources. Al chatbots, conversation scanning tools and dynam¬ ic content creation tools will require systems that un¬ derstand the semantics and intent of people when they communicate. This requires research into ma¬ chine learning and natural language processing tools and the creation of an ontology for foreign affairs topics. Essentially, the Al system will need to learn the language of foreign affairs. This requires a signif¬ icant, long-term investment of resources. However, such a system will have additional diplomatic ben¬ efits outside of public diplomacy. Al tools to extract meaning from conversations could be used to discov¬ er hidden knowledge and patterns from internal de¬ partment communications. Al scanning tools could be used as supplementary analytical “assistants” for ACPD I Understanding the Challenges of A! and Computational Propaganda to PD 45 U.S. diplomats. These would increase employee situa¬ tional awareness, productivity and efficiency through enhanced predictive abilities and automated coun¬ try-specific event monitoring. Al content creation tools would enable semi-automated speech-writing and document preparation. Al conversational tools could be used to share information among the De¬ partment’s various networks of contacts to support diplomatic initiatives.26 The Department should also build capabilities for personalized targeting of persuasive communica¬ tions based on individual psychological profiling and big-data analytics (within the parameters of key statutes like the Privacy Act of 1974 and Smith-Mundt The U.S. government must adapt to the new threats from Al-enabled computational propaganda before the online information environment is toxified from machine-generated speech. At the turn of the 21st century, the internet was seen as a mortal threat to authoritarian regimes due to increased openness, access to information and ability to organize online. Some regimes neutralized this threat and a number of actors have turned our own technology against us by hacking free speech. MADCOMs have given foreign actors the ability to reach directly into the United States with unprecedented reach, speed and effectiveness. Artificial intelligence technologies will soon boost those capabilities with additional velocity, “ The machines are here and they want to have a word with us. ” Act). This would allow personalized messaging, or mass-market communications campaigns to accu¬ rately tailor messages to target populations. These tools would include affective computing technolo¬ gies to identify and convey the appropriate emotional tone in messaging. Machine-learning tools can also be used to run test campaigns on individuals that help optimize mass market campaigns for similar au¬ diences. Public diplomacy business processes will need to adapt to use the new tools and operate closer to machine speed. Content creation processes must accelerate and approval cycles must compress. The Department will need to gain comfort with machines operating autonomously without tedious clearance processes. The Department will also need to consid¬ er forming permanent rapid-response task forces that can respond to emerging computational propa¬ ganda campaigns with high velocity rather than rely¬ ing on one-off efforts. The Department should consider how to promote the development and availability of real-time fact-checking27 and bot detection tools. These may be useful for validating news and content and identi¬ fying whether an account is machine-driven. Stand¬ alone tools would rely on the user proactively access¬ ing them, but technology companies could integrate these functions into browsers, applications and oth¬ er platforms. personalized targeting, human mimicry, increased operational tempo and machine learning. Regulation is one option,28 but hardly the only option. The gov¬ ernment needs a comprehensive, strategic response to the current threat but also to the environment we are rapidly moving into. Responding to the threat of Al-enabled machine-driv¬ en communications tools, and capitalizing on op¬ portunities posed by Al MADCOMs, will require a conceptual leap in understanding into the world of personalized, psychology-based, machine-driven persuasion. This struggle will take place 24 hours a day, every day and will become so rapid and com¬ plex that humans cannot hope to operate effectively alone. Rather than using new tools to do the same old processes better (e.g., dynamically generated press releases), new artificial intelligence and MADCOM tools must be integrated thoroughly into PD business processes that are reinvented and reimagined to take advantage of those tools. This will require PD profes¬ sionals to learn to team with intelligent machines. It will also require Al tools of high enough quality that we can trust them with the autonomy they will re¬ quire. This will take years of hard work to accomplish. We must accelerate our efforts immediately. The machines are here and they wantto havea word with us. Our level of preparation forthis emerging re¬ ality will determine the fate of the internet, our soci- etyandourdemocracy. 46 ACPD I Can Public Diplomacy Survive the Internet? ENDNOTES 1. See Meet Xiaolce, Cortana’s Little Sister: https://blogs.bing.com/search/2014/09/Q5/meet-xiaoice-cortanas-little-sister/; and https://en.wikipedia.org/wiki/Xiaoice 2. MADCOMs can include anything from robo-dialing telemarketing tools, to Al home assistants like Amazon Echo, to cutting edge (and mysterious) Als like “T” that dynamically create thousands of fake news videos for YouTube. 3. See this video from DARPA that explains how Al evolved and how machine learning segments manifold data: https://www.you- tube.com/watch?v=-OQ1 G3tSYpU 4. See For Sympathetic Ear, More Chinese Turn to Smartphone Program: https://nyti.ms/2peM3T6; and Meet Xiaolce, Cortana’s Little Sister: https://blogs.bing.com/search/2014/09/05/meet-xiaoice-cortanas-little-sister/ 5. See AP’s ‘robot journalists’ are writing their own stories now: http://www.theverge.com/201 5/1 /29/7939067/ap-journalism-au- tomation-robots-financial-reporting 6. See An Al Written Novel Has Passed Literary Prize Screening: https://futurism.com/this-ai-wrote-a-novel-and-the-work-passed- the-first-round-of-a-national-literary-award/ 7. See IBM Debating Technologies: http://researcher.watson.ibm.com/researcher/view_group.php?id=5443 8. See Affective Computing: https://en.wikipedia.org/wiki/Affective_computing 9. See Research on Affective Pattern Recognition and Modeling: http://affect.media.mit.ed u/areas.php?id=recognizing 10. See This Freaky Baby Could Be the Future of Al : https://www.youtube.com/watch?v=yzFW4-dvFDA&feature=youtu.be 11. See Face2Face: Real-time Face Capture and Reenactment of RGB Videos: http://www.graphics.stanford.edu/~niessner/thies- 2016face.html 12. See Face2Face: Real-time Face Capture and Reenactment of RGB Videos: https://www.youtube.com/watch?v=ohmajJTcpNk&- feature=youtu.be 13. See Concatenation synthesis https://en.wikipedia.Org/wiki/Speech_synthesis#Concatenation_synthesis 14. See A system for voice conversion based on probabilistic classification and a harmonic plus noise model: http://ieeexplore.ieee. org/document/674422/?reload=true 15. See Google’s DeepMind develops creepy, ultra-realistic human speech synthesis: www.geek.com/tech/googles-deepmind-devel- ops-creepy-ultra-realistic-human-speech-synthesis-1 670362/ 16. See The Secretive World of Selling Data About You: http://www.newsweek.com/secretive-world-selling-data-about-you-464789 17. See The Secret Agenda of a Facebook Quiz: https://www.nytimes.com/2016/11/20/opinion/the-secret-agenda-of-a-facebook- quiz.html 18. See Behind the scenes at Donald Trump’s UK digital war room: http://news.sky.com/story/behind-the-scenes-at-donald-trumps- uk-digital-war-room-1 0626155 19. Have you ever taken a personality test on Facebook? If so, you’ve probably given a marketer your personality and possibly psy¬ chological profile, along with your name, email address and friend list. 20. See the accompanying paper “Understanding the Psychology Behind Computational Propaganda” 21. This balance between MADCOMs precisely targeted towards people and MADCOMs targeted towards machines and people is an unknown variable. Raising the costs to MADCOMs through filtering might have a significant positive impact on the information environment. As an analogy, filtering reduces but does not eliminate email spam. 22. Facebook’s ‘Real Name’ policy and Twitter’s automation policy are two examples of efforts to fight bots and fake accounts. 23. Politicalbots.org is an excellent resource for computational propaganda information and has a recommended reading list. The Observatory on Social Media publishes research on information diffusion in social media. The author maintains a directory of Al policy issues and resources on Medium. 24. The State Department already uses MADCOMs and Al as they are built into social media platforms; e.g. Facebook targeted adver¬ tising relies on machine learning to reach the desired audience. 25. See the accompanying paper: “Understanding the Psychology Behind Computational Propaganda” ACPD I Understanding the Challenges of A! and Computational Propaganda to PD 47 26. Creation of this initial ontological system would likely cost upwards of $5 million and would require significant annual invest¬ ments to build capabilities and expertise in specific topics. However, the benefits in terms of increased capabilities, productivity and efficiency argue for this long-term investment. Otherwise the State Department will soon find that it is the only comparably sized, global organization without enterprise Al tools built into its business processes, and will suffer for this lack of capability. 27. See Automated Fact Checking: The Holy Grail of Political Communication: http://nordicapis.com/automated-fact-check- ing- the-holy-grail-of-political-communication/ 28. This raises 1st Amendment Concerns, but foreign actors do not have Constitutional protections, nor do machines. The United States has regulated false speech in areas where there is consensus that the false speech is against the public interest, like false advertising, slander and libel cases. 4 | 4*j| vj wa jpn j Jgjp 1 i i ; T B, 8 " Wj f " J ^ ml ■5rj*A.|i 1 V V H H * &■ ^ r. .■> 1 -Bj ■ Vi y T1 P j , * " 1 u H| i 9 ■! , jj 1 Ll T 1 I ■ < & PSYCHOLOGICAL PRINCIPLES FOR PUBLIC DIPLOMACY IN AN EVOLVING INFORMATION ECOSYSTEM By Jeffrey T. Hancock, Professor of Communication, Stanford University In this essay I focus on some of the psychological aspects of how communication technology affects the way that people deceive and trust one another. The deep concerns we’ve been facing lately about a “post-truth society” are really a reflection of how we can trust one another in a world dominated by social media, a place in which people we may or may not know can communicate with us at any time and from anywhere. How can we tell if someone is lying to us in their tweet, their Facebook post, the news that they shared via a text or an online video? Worse yet, some of those people might not even be people, but bots coordinating to promote some propaganda or com¬ mercial interest. Concerns about misinformation, fake news and whether my new friend is a bot can lead us to con¬ clude that social media is dramatically increasing deception in the world, and that soon we’ll be unable to trust one another, or establish what information is true. But as Ambassador Bruce Wharton, Acting Un¬ der Secretary for Public Diplomacy and Public Affairs, suggests in this report, I believe that we are not in a post-truth society. Although we’re paying more atten¬ tion to the topic of truth and evidence, social media and related technologies do not spell the end of hon¬ esty. Instead, that technology will transform how de¬ ception takes place, how we detect lies and how we come to trust one another. These changes to truth-telling and trust will not be random or unpredictable, but will be driven by prin¬ ciples and factors that the social sciences have been identifying over the past century. We need not throw out the book of psychology, for example, to under¬ stand how public diplomacy needs to adapt to the changes wrought by social media. Below I review these principles and provide an overview of the latest research on deception detection and trust, conclud¬ ing with insights on what those engaged in public di¬ plomacy need to address most to succeed in an evolv¬ ing communication and information environment. AN ANCIENT PSYCHOLOGICAL CONCERN Every generation tends to think that the current gen¬ eration is less honest than the previous generation. This is an old human concern. In western culture we have Diogenes, the Greek philosopher who searched for a single honest man, failing to ever find one. In the east, the Chinese were so concerned with hon¬ esty that they developed the first deception detection technique over 2000 years ago. They put dry rice into a suspected liar’s mouth. If the suspect couldn’t talk then it suggested they didn’t have enough saliva, which the Chinese believed was a symptom of lying. These ancient examples highlight that it is important to historicize our current concern with deception and misinformation with social media. As a society, we have long been concerned with truthfulness, and this concern is often made more salient when new communication technologies are introduced, from the printing press to the radio. With social media, we see a similar pattern. To put deception and social me¬ dia into context, it is useful to understand what psy¬ chology has uncovered about how deception works. A PRIMER ON DECEPTION DETECTION What does the hundreds of studies by psychologists and communication researchers on deception detec¬ tion tell us? First, and surprisingly, there is no reliable cue that always indicates whether a person is lying. There is no Pinocchio’s nose, as much as TV shows or self-help magazines would like us to believe. While there are some contexts, such as interrogations and interviews and other high-stake situations, where reliable cues to deception can be elicited, there is no nonverbal cue that reveals lying in all the different domains of human discourse. Recent work suggests that deception can be more reliably revealed in lan¬ guage patterns, but sophisticated training or com¬ puter programs are required to identify the linguistic footprints of lies. 50 ACPD I Can Public Diplomacy Survive the Internet? An overall meta-analysis of hundreds of deception experiments reveals that humans perform at chance levels (54 percent) when detecting deception. We really aren’t very good at telling if someone is lying based on verbal or nonverbal cues, in part because there are no reliable cues. This difficulty in detecting deception transfers over to social media deception. In studies examining phishing attacks, where decep¬ tive emails are used to access sensitive information (as was the case in the hacking of the Clinton cam¬ paign accounts), even sophisticated users can be de¬ ceived, online or off. In fact, there is only one reliable finding in every de¬ ception detection study: people tend to trust what others say, an effect called the truth bias. Our default state is to trust what other people say. This bias is deception by just relying on cues in the message, we now have information tools available that can help investigate potential deceptions that were hard to imagine just a few years ago (for more on the value of fact-checking efforts, see Ethan Porter’s essay in this report). The last finding from the deception literature that is important to share is that people lie for a reason, and these reasons are widely varied. While this may seem obvious, it is important to note this when considering how deception operates in the current environment. Fake news articles in the last election cycle were pro¬ duced sometimes to influence voters, but more often the motives were simply profit. Without considering the reasons for deception, it is impossible to counter them. “Every generation tends to think that the cur¬ rent generation is iess honestthan the previous generation. This is an old human concern. ” actually quite rational — most of the messages that a person encounters in a day are honest, so being bi¬ ased toward the truth is almost always the correct response. This tendency to trust messages is, of course, one of the reasons that lies can succeed, but it’s important to note how fundamental the truth bias is. Language philosophers even argue that for lan¬ guage to work we must assume a cooperative part¬ ner, suggesting that the truth bias is fundamental to communication. There are two other important findings from the de¬ ception literature that are relevant to public diploma¬ cy. While we have a difficult time detecting deception from cues, likeeyegazeorvocal pitch, peoplecan de- tectlieswhentheyhavethetime,resourcesand moti¬ vation. Lies are often discovered through contradict¬ ing information from a third source, or evidence that challenges a deceptive account. Much like the way police officers investigate witness statements and suspectalibis, people often determinethatthey have been deceived by seeking outand learning from oth¬ er information. Our recent work suggests that most lies that people detect rely on information from oth¬ ers, or come from using search engines like Google orexamining activities posted on social network sites like Facebook. Thus, while we may be bad at detecting DECEPTION AND TRUST IN AN EVOLV- INGCOMMUNICATION ENVIRON¬ MENT Deception is the deliberate attempt to create a false belief in another, so to understand deception it is important to understand how beliefs are formed. Cognitive science research reveals that we tend to believe information that we receive, which gives rise to the truth bias described above. This initial belief is “sticky” and the belief can persist even when it is later shown to be false or incorrect, suggesting that the first mover in sharing information has an advan¬ tage. Further, as information is repeated, these initial beliefs become even more difficult to overturn, a fact that advertisers know well. At the same time people are also continuously evalu¬ ating the validity of their understanding of the world. This process is called “epistemic vigilance,” a contin¬ uous process checking that the information that a person believes they know about the world is accu¬ rate. Epistemic vigilance works in parallel with the truth bias, alert to any signals that information about the world may be incorrect, such as inconsistencies ACPDj 51 across sources. Thus, while our default is to trust in¬ coming information, people also evaluate their infor¬ mation environmentto ensure that their understand¬ ing of their world is valid. As our information ecology evolves to be more me¬ diatized1 and digital, the operation of epistemicvigi- lanceneedstoevolveaswell.Considerthesharingof a fake news story by Shawn on Facebook that is read by Markos. There are several signals Markos’ epis- temic vigilance might rely on: the degree to which Markos knows and trusts Shawn, how often the ar¬ ticle has been “liked” by others, and the number of times the article has been shared. If these signals are all high, then there is little to trigger additional vigilance or signal for Markos to move awayfrom his default state of trust. Since the alarms over fake news emerged, social network sites and journalists have begun to develop additional signals that can help people assess the validity of information. For example, Facebook now allows users to flag stories that may be fake. These stories are then examined by fact checkers, and if the story is fake an alert is shown whenever the story is shared on the platform. This kind of signal functions to trigger epistemic vigilance and help individuals make decisions about whether to trust the informa¬ tion or not. While this is a promising development, it is still too early to measure the effectiveness of these measures. This change nonetheless points to at least one direction for improving our ability to assess in¬ formation in social media. Much more is required to establish trustworthy com¬ munication in our evolving information environment. While substantial work is required to continuing im¬ proving the trustworthiness of our cyber systems, we also need to better understand the social aspects of these newtechnologies. In ourown work, we have begun to look at how people reason about social technologies, like Facebook or Twitter newsfeeds. We find that people have folk theories about these tech¬ nologies, which represent the person’s general un¬ derstanding of how a system works. These theories weigh questions, such as: Howdo algorithms decide to share info? How are sources perceived? To what degree do these systems validate information? For example, we find that some people think of Face- book’s newsfeed as a personal shopper, helping the person find things of interest to them. Others, how¬ ever, think of the newsfeed as a spy or as paparaz¬ zi, concerned that the system is designed to exploit them for the gain of others. Without knowing more about people’s folk theories of these complex sys¬ tems, it is difficult to predict how audiences will react to messages that are shared through them, such as whether they trust them or not. SOME REASONS FOR HOPE: LESSONS FROM THE SHARING ECONOMY There is substantial reason to be optimistic in the long term about truth and trust with technology. Although trust in institutions, such as media, government and religion, has been in decline for over a decade, there has been substantial trust observed in how people are believing each other via technology. For example, when purchasing new products and services, most people will rely on online reviews to make decisions about what hotel to reserve or which car to buy. More people trust peers when making these decisions than any other form of media. In another domain, we see huge trust in social support groups that operate on the internet, with strangers providing support and advice to other strangers, trusting one another to help each other face cancer, overcome the loss of loved ones, or how to recover from bankruptcy. This inversion of trust, decreasing trust in institu¬ tions but rise in interpersonal trust, can also be ob¬ served in the sharing economy, from home-sharing to car-sharing. Considerthe level of trust required to allow strangers to stay in your home. Or the amount of trust required to hop into a stranger’s car late at night in a strange neighborhood. How does trust op¬ erate in this multi-billion-dollar economy, and what insights can it provide for public diplomacy in this evolving communication environment? First, the trust placed in these services is warranted. Very few rides on Lyft or Uber result in any negative incident. The same is true for house-sharing services like Airbnb. One reason for this is that the users’ goals are aligned. One user would like to sell their service while the other user wants to buy this service. When goals are aligned, trust can facilitate many social transactions. Public diplomats know this well — mes¬ sages must be aligned with the goals of the audience or the partner. Forgetting this can undermine any diplomatic enterprise. Second, while we usually think about the person tak¬ ing the risk when we think about trust situations, it is important to consider the psychological dynamics on the other side of the risk, the person being trusted. 52 ACPD I Can Public Diplomacy Survive the Internet? When people are trusted with something valuable, such as being allowed into a stranger’s home, they of¬ ten experience feelings of responsibility and are even nervous about harming the other person. Indeed, many Airbnb hosts report that their homes are in great shape after renting them out. Trust often leads to trustworthy behavior. Third, users of these services believe that there is in¬ frastructure in place to protect them from violations of trust. Users expect that brands like Airbnb will re¬ imburse them for any damages. Further, there trust is built on layers of older, legacy infrastructure, like law enforcement and financial regulation. In addition to the brand of a service, such as Airbnb, users expect to be supported by the enforcement of legal institutions put in place long before social media came on the scene, from the police department to the Consumer Financial Protection Bureau. These layers of infra¬ structure that build trust should also be considered in public diplomacy campaigns. Programming that aims to establish relationships, and build on those re¬ lationships over time, is likely to result in robust net¬ works of shared interests and understanding. Finally, technology plays an important role. Users of Uber report feeling safe in part because the app constantly records where they are. They believe that should something go wrong, there will be a re¬ cord, and that this record keeps people honest. In¬ deed, one of the most important transformations of the communication environment is the record of behaviors, taking millions of rides with strangers and allowing millions of strangers to stay at their homes. LESSONS FOR PUBLIC DIPLOMACY: TRAINING AND IMPLEMENTATION As people engaged in public diplomacy adapt to the evolving communication environment to engage with foreign audiences, what are some of the keys for success? The first is recognizing that the goals and values of the United States and its allies and adver¬ saries are paramount. It is important to keep a focus on goals, objectives and our own values. Deception is often detrimental in the long term, and the costs to reputations can be severe. In one study asking peo¬ ple to rank traits, the one ranked lowest from a total of over 500 was “liar.” It is also important for those serving in a public di¬ plomacy role to receive new forms of training and education. This training should involve an emphasis on media literacy, including both the social science of technology and also enhancing technical skills. One model may be the computational journalism pro¬ gram at Stanford University, which seeks to trans¬ form journalism by providing journalists with com¬ putational capacities that will change how they can investigate issues of public interest. I can imagine a new program for “computational diplomacy” that has similar goals, to develop new skills for diplomacy that incorporate computational abilities and social sci- “Trust often leads to trustworthy behavior” so many more behaviors and actions. Deception is made more difficult when there’s a record, as most politicians caught up in personal scandals can attest. Public diplomacy practitioners must remain vigilant in ensuring their outreach is transparent and mes¬ saging grounded in empiricism. Overall, the sharing economy points out a possible path for how trust can emerge and thrive in a fully mediatized information environment, and how many partners and layers of infrastructure can be lever¬ aged to make it work. Importantly, it is also reassur¬ ing to see that people trust one another with their ence training for understanding the social and tech¬ nical aspects of new communication environments. Finally, the United States needs to continue to devel¬ op its technical capabilities to be able to detect and counter misinformation and other attacks by hostile others. Importantly, this should be coupled with a similar investment in training in the social sciences, from the psychology of technology discussed here to social network analysis. All of these technological ca¬ pacities and social analysis skills will be required to best engage our foreign audiences. ACPDj 53 ENDNOTES 1. In media studies, mediatization is a theory suggesting that the media shapes and frames the processes and dis¬ course of political communication as well as the society in which that communication takes place. Image: CC BY-NC-ND 2.0, Martin Sommer FACTS MATTER, AND PEOPLE CARE AN EMPIRICAL PERSPECTIVE By Ethan Porter, Asst. Professorat George Washington University School of Media and Public Affairs Does fact-checking work? That is, if we provide peo¬ ple with information that corrects their mispercep¬ tions, do they respond by accepting the new infor¬ mation — or do they reject it and cling to their prior beliefs, no matter how wrong they might be? There is ample reason to be pessimistic. Hardly a day goes by without a pundit or public figure bemoaning the arriv¬ al of a “post-truth” age, in which the stuff of facts and evidence no longer matters. Although media institu¬ tions devote considerable resources to fact-check¬ ing, the naked eye suggests that such efforts are mostly for naught. One only has to spend a little time on the internet, or talk with family members, to be¬ come acquainted with a vast variety of confidently asserted mistruths offered by all sides of the political spectrum. Researchers are of two camps on this matter. On the one hand, some have offered evidence implying that efforts to raise the level of political knowledge and correct misinformation are unlikely to succeed. According to this school of thought, people know very little about politics and they are so committed to their political beliefs that they have difficulty ac¬ cepting facts that challenge those beliefs.1 In fact, at¬ tempting to correct misperceptions may only serve to strengthen people’s commitments to their misper¬ ceptions.2 According to another perspective, howev¬ er, pessimism about citizens’ political knowledge and their receptivity toward factual information is grossly overstated. Though people may not know as much about politics as policymakers and educators might wish, they are quite capable of learning more, even when doing so requires them to break from their po¬ litical commitments or to think through complicated policy issues.3 In this essay, I review both perspectives. I then de¬ scribe three studies, administered over large num¬ bers of people in the United States and the United Kingdom, that puncture the pessimists’ consensus. The studies make clear that, while people may not know much about politics, they can learn more and, crucially, that their misperceptions can be corrected. Fact-checking may not work to the degree that some hope it will but, as a general matter, it does seem to work. Not only can people learn about politics, but they can do so even when the issues are complicated and when the facts challenge their most cherished political beliefs. Academic research offers many reasons to be skepti¬ cal that citizens can learn and that fact-checking can work. From a bird’s eye view, the public appears to be terribly misinformed. For example, in both the Unit¬ ed States and the United Kingdom — two of the most well-educated states in the world — citizens dramat¬ ically over-estimate the amount of money their gov¬ ernments spend on foreign aid, sometimes by factors of ten.4 The picture only gets worse from here. Not only do citizens know very little about politics, they are virtually incapable of learning more. Their level of commitment to their party of choice is so strong that their partisanship effectively precludes them from learning new information that conflicts with their political commitments. Partisanship, it has been said, instills what has been evocatively called a “perceptu¬ al screen” that comes between everyday people and the empirical world around them. When confronted with information that challenges their political be¬ liefs, they may aggressively seek out information that confirms those beliefs.5 Perhaps the most infamous example of citizens’ un¬ willingness to brook new, challenging facts comes in the form of the supposed “backfire effect.” The back¬ fire effect proposes that, when people are presented with facts that correct misstatements made by their co-partisans, they will become more convinced of the misstatements. That is, rather than move toward the factually accurate position, they will move in the other direction. First identified in the study of attitudes to¬ ward the Iraq War, during which time conservatives backfired against corrections relating to the United States failure to find weapons of mass destruction (WMD), the backfire effect has also been observed in relationship to attitudes about tax policy and vac¬ cinations.6 As the name of the phenomenon implies, fact-checking does not simply fail — it backfires, pro¬ ducing the opposite effect of what those who imple¬ ment it intend. For nearly just as long as some scholars have offered reasons to think that citizens cannot learn, others 56 ACPD I Can Public Diplomacy Survive the Internet? have disagreed. Several have argued that well-known estimations of political ignorance are wildly overstat¬ ed.7 Others have shown that offering small monetary incentives for accurate responses can dramatically reduce the amount of partisan bias that colors an¬ swers to factual questions.8 Still others have shown that citizens can, over time and with some effort, in¬ crease their store of political knowledge — they can “learn together, slowly.”9 Three recent studies have tested these competing schools of thought. In the first study, conducted in Spring 2016, we attempted to map out the backfire effect among U.S. citizens, to identify the specific pol¬ icy areas that would provoke certain ideological and partisan groups to backfire. Similar to the original studies that found backfire in relationship to WMDs in Iraq, we searched for instances in which political office holders or political office seekers made factu¬ al misstatements. We then randomly exposed some survey takers to a correction to the misstatement.10 In all cases, the correction explicitly referred to neu¬ tral government data. We then asked all subjects whether they agreed with the factually accurate po¬ sition espoused by the politician who had made the statement.11 We were careful to include a broad variety of issues and politicians from both parties. (Perhaps unsur¬ prisingly, we found many examples of misstatements from both parties.) In one experiment, subjects were randomly assigned to see only a misstatement and then a neutral correction. For example, all respon¬ dents were presented with the following statement by Hillary Clinton: ”We need to get back into the habit of ac¬ tually rewarding workers with increases in their paychecks... Warren Buffett has said it, but so have a lot of other people. There’s something wrong when hedge fund man¬ agers make more, and pay less in taxes, than nurses or truck drivers.” Then, those randomly assigned to see the correction saw: “In fact, according to the Bureau of Labor Statistics, the average hedge fund manag¬ er pays about 20 times as much income tax as the average truck driver or nurse.” All subjects were then asked to agree or disagree with the factually incorrect statement offered by the politician. We were not afraid to identify Hillary Clin¬ ton’s partisan affiliation; we placed it squarely next to her name. By the logic of backfire, this presentation should compel Clinton’s co-partisans to reject the correction provided, and become more convinced of the factually inaccurate position articulated by Clin¬ ton. We also found instances of policy areas in which members of both parties had made misstatements. For example, Democrats and Republicans have ex¬ aggerated the amount of U.S. debt that China owns. Again, we provided all subjects with these biparti¬ san misperceptions, randomly showed some survey takers a correction based on data from a neutral government source and then asked everyone if they agreed or disagreed with the misperception. Finally, because some of the earlier backfire work embedded misstatements and the attendant corrections in (fic¬ titious) newspaper articles, we created newspaper articles of our own. We took a new set of misstate¬ ments made by leaders of both parties, crafted arti¬ cles around such misstatements and, in some of the articles, included corrections to the misstatements based on neutral government data. “...when the correction went against their co-partisan, they sided with the correction over and above their partisanship. When a feilow partisan is being corrected, people still learn from the correction, albeit grudgingly. ” ACPD I Facts Matter, and People Care: An Empirical Perspective 57 All told, this study enrolled 8, 1 00 people and tested 36 different issues’ capacity to generate backfire. On 35 issues, we observed no backfire. Regardless of their own partisan beliefs, and regardless of the party affil¬ iation of the politician being corrected, those survey takers who saw the correction became more con¬ vinced of the factually accurate position. To be sure, people were more reluctant to accept the facts when the facts corrected a fellow partisan than when the facts corrected a member of the opposite party. But even when the correction went against their co-par¬ tisan, they sided with the correction over and above their partisanship. When a fellow partisan is being corrected, people still learn from the correction, al¬ beit grudgingly. As noted, we observed one instance of backfire. In this case, it was the same issue that yielded back¬ fire in the initial backfire study — whether the United States had found WMD in Iraq. Once again, conser¬ vatives shown a correction clarifying that no WMD were in fact found became more convinced that WMD were found. However, even this replicable instance of backfire was quickly overturned when we changed the wording of the survey question. When the ques¬ tion was made more succinct, backfire vanished. Once again, regardless of their partisan identifica¬ tion, everyone was willing to accept the factual cor¬ rection provided to them.12 Of the 36 issues we tested, 35 did not generate back¬ fire; and the one that did was highly susceptible to question-wording effects. Just as important as what we found — that backfire is, at best, a byproduct of question wording — was what we didn’t find. We found no evidence that people exposed to more factual cor¬ rections over the course of our studies became more or less willing to accept factual corrections. Nor did we find evidence that the order of factual corrections one saw had any impact on one’s willingness to indi¬ cate agreement with the facts. In attempting to map out the contours of the backfire effect, we found it was far smaller than previous research indicated. In the study just described, all experiments were conducted in the United States in Spring 2016 — as the presidential race was ongoing, but before both parties had selected their nominees. According to the pessimistic school of thought, the relationship between facts and partisan loyalty matters quite a bit. And if this is the case, then it stands to reason that the relationship might be at its apex after both parties have selected nominees. To account for this possibility, we conducted an additional study during the general election. This study consisted of two ex¬ periments designed to whether the white heat of the general election would make people more suscepti¬ ble to backfire than they would be otherwise.13 In the first experiment, we exploited a statement made by then-candidate Trump during the 2016 Re¬ publican National Convention about crime. Not only was the statement at odds with data produced by the Federal Bureau of Investigation, but when media fig¬ ures critiqued him for it, his campaign responded by denigrating the quality of the data and implying the existence of an anti-Trump conspiracy within the FBI. Once again, we devised several fictitious news arti¬ cles. In one version, we merely relayed Trump’s state¬ ments. In another, we included a correction. In still another, we included the correction and the Trump campaign’s denigration of the data. In yet another, we included the correction, the Trump campaign’s denigration of the data and its insinuation that an an¬ ti-Trump conspiracy within the FBI had meddled with the data. In the second experiment, we exploited a misstate¬ ment made by Trump during the first debate of the general election. While we conducted the experi¬ ment on the convention statements six weeks after the convention, we conducted the experiment about the debate on the same night as the debate. During the debate, he made a claim about unemployment that diverged from available Bureau of Labor Statis¬ tics data. On the night of the debate, we paid people to watch the debate. After it ended, we showed ev¬ eryone Trump’s inaccurate unemployment claim and then showed some the BLS correction. In both experiments, all subjects proved willing to ac¬ cept the factual correction provided to them.14 Trump supporters did the same, even though their candi¬ date was being corrected in the midst of a presiden¬ tial election. However, their views of their preferred candidate did not budge. In the convention study, Trump supporters who saw the correction became no less favorable to Donald Trump — even though they accepted the content of the correction. We ob¬ served a similar pattern in the debate study. Trump supporters who saw the correction did not subse¬ quently change their minds about candidate Trump. Though they accepted the factual correction, thereby conceding that their preferred candidate had made statements at odds with the facts, they were no less willing to support him as a result. The verdict is clear: People can accept factual correction, even when do¬ ing so forces them to break from their preferred po- 58 ACPD I Can Public Diplomacy Survive the Internet? “To issue a correction does not amount to putting the thumb on the scale and fa voring one candidate o ver another... the only consequence of fact-checking... is that it increases the extent to which the public be¬ lieves in factually accurate information. ” litical parties, and even in the midst of a presidential election. Can people also learn about politics when a complex policy matter, independent of partisan politics, is at stake? A third study suggests that they can. In 2014, the United Kingdom mailed “taxpayer receipts” that offered itemized descriptions of how government had spent the tax money it collected, presented on a per-capita basis, to 26 million taxpayers. Work¬ ing in coordination with the U.K. tax authorities, we empaneled a large group of survey respondents.15 We then randomly assigned some people to receive reminders and encouragements about the taxpay¬ er receipt that they would receive soon in the mail. We surveyed everyone before and after the receipts went out about their level of political knowledge and a host of political attitudes. To measure levels of po¬ litical knowledge, we asked subjects to estimate how much their government had spent on various items over the previous year. This information, contained in the receipts themselves, has confounded many people in many countries. As mentioned above, peo¬ ple are notoriously lacking in knowledge about how much their governments spend on foreign aid — and we asked everyone precisely this question.16 Across a range of knowledge measures, we found that the receipts caused a significant uptick in politi¬ cal knowledge. Measured a number of different ways, we find that, indeed, the receipts made people more likely to provide accurate estimates of the amount their government was spending on foreign aid. We found similar effects for other uses of government money. However, we found no effects on related po¬ litical attitudes. In this study, people were not asked to break from a fellow partisan and accept a factual correction; instead, they were asked to break from their own prior misperceptions. And they did so. To be sure, they did not know much about politics before the receipts went out. But they could learn. And they could do so without changing their views on related matters. For those with public responsibilities, the erosion of the pessimistic consensus has significant im¬ plications. Fact-checkers should take their foot off the break. Where they see erroneous claims, they should be unafraid to intervene. When administered to counter a fiction, a correction can prevail even over the power of partisanship. This is true not only when political figures make misstatements, but when vast numbers of citizens believe in policy mistruths, as with foreign aid. Again and again, the facts get through. That we find no evidence that increasing political knowledge causes related attitude change should further fuel the fire of those who wish to increase political knowledge. To issue a correction does not amount to putting the thumb on the scale and fa¬ voring one candidate over another. Instead, the only consequence of fact-checking that we can consis¬ tently detect is that it increases the extent to which the public believes in factually accurate information. Whatever one’s political position, this seems a laud- ableend. ACPD I Facts Matter, and People Care: An Empirical Perspective 59 ENDNOTES 1. Carpini, Michael Delli and Scott Keeter. 1996. What Americans Know About Politics and Why It Matters. New Haven, Connecticut: Yale University Press. 2. Nyhan, Brendan and Jason Reifler. 2010. "When Corrections Fail.” Political Behavior. 32 (2): 303-330. 3. Hill, Seth. 2017. "Learning Together, Slowly.” Journal of Politics. Forthcoming. 4. Klein, Ezra. 2013. “The Budget Myth That Won’t Die.” Washington Post Wonkblog, https://www.washingtonpost.com/ news/wonk/wp/2013/1 1/07/the-budget-myth-that-justwont-die-americans-still-think-28-percent-of-the-budget- goes-to-foreign-aid/; Kings College London and Ipsos Mori. 201 3. “Perils of Perception.” Survey Release, 9 July 2013. https://www.ipsos-mori.com/researchpublications/researcharchive/3188/Perceptions-are-not-reality-the-top-10- we-get-wrong . as px#_ed n 6 . 5. Lodge, Milton and Charles Taber. 201 3. The Rationalizing Voter. New York, New York: Cambridge University Press. 6. Nyhan and Reifler 2010; Flynn, D.J., Brendan Nyhan and Jason Reifler. 2016. "The Nature and Origins of Mispercep¬ tions: Understanding False and Unsupported Beliefs about Politics.” Advances in Political Psychology. 7. Gibson, James L. and Gregory Caldiera. 2009. Citizens, Courts and Confirmations. Princeton, New Jersey: Princeton University Press. 8. Bullock, John G, Alan S. Gerber, Seth J. Hill, and Gregory A. Huber. 2015. “Partisan Bias in Factual Beliefs about Poli¬ tics.” Quarterly Journal of Political Science. 10: 519-78. 9. Hill 2017. 10. In the first study described, all subjects were recruited over Amazon’s Mechanical Turk, a low-cost platform survey subject acquisition in wide use in the social sciences (Berinsky, Huber and Lenz 2012). 11. Wood, Thomas J. and Ethan Porter. 2016. "The Elusive Backfire Effect.” Wood, Thomas and Porter, Ethan, The Elu¬ sive Backfire Effect: Mass Attitudes’ Steadfast Factual Adherence (August 5, 2016). Available at SSRN: https://ssrn. com/abstract=281 9073 12. The original question wording read: "Immediately before the U.S. invasion, Iraq had an active weapons of mass destruction program, the ability to produce these weapons, and large stockpiles of WMD, but Saddam Hussein was able to hide or destroy these weapons right before U.S. forces arrived.” We changed that to: "Following the U.S. invasion of Iraq in 2003, U.S. forces did not find weapons of mass destruction.” 13. Nyhan, Brendan. November 5, 2016. "Fact-Checking Can Change Views? We Rate That As Mostly True.” The New YorkTimes. 14. For the experiment about Republican National Convention, we recruited subjects from Mechanical Turk and a na¬ tionally representative sample from Morning Consult. Results were indistinguishable across both samples. For the experiment about the debate, we used Mechanical Turk to recruit subjects. 15. We used YouGov U.K. for this study, which provided us with a nationally representative sample. 16. Barnes, Lucy and Feller, Avi and Haselswerdt, Jake and Porter, Ethan, Can Citizens Learn? An Evaluation of the Tax¬ payer Receipt (June 6, 201 6). Available at SSRN : https://ssrn.com/abstract=2877248 Voice of America VOA: A WEAPON OF TRUTH IN THE WAR OF WORDS By Amanda Bennett, Director, Voice of America Does truth matter? In a simple word, yes. Are we liv¬ ing in a post-truth era? No. Despite the flood of misin¬ formation and our own fears of being overcome by it, history shows that truth is one of our most powerful weapons. Against the current backdrop of propa¬ ganda and falsified news in the global media environ¬ ment, truth and facts stand out — informing, educat¬ ing and empowering citizenry. Voice of America, for 75 years, has been offering just that — the truth. In its first broadcast, William Har¬ lan Hale said " The news may be good for us. The news may be bad. But we shall tell you the truth." Much has changed since the days of Nazi propaganda and shortwave radio. The revolution in technology and vast proliferation in social media use over just the past decade has dramatically upended the way infor¬ mation is gathered and shared. These technological advancements initially raised hopes for more informed and connected societ¬ ies — and perhaps a diminishing need for a govern¬ ment-funded broadcaster like VOA. However, this optimism must be tempered by today’s reality. State and non-state actors alike are trying to undermine Western democratic principles and are using tradi¬ tional and social media to distribute their message. Bots and trolls disrupt civil debate online and web¬ sites masquerading as news sites offer sensational headlines just to make a quick, click-through, dollar. The effort to manipulate or incite is not new. In the past, whole swathes of the globe were denied access to truthful information. However, the problem today is compounded by the ease and speed with which false information can be disseminated. This can cre¬ ate a sense of chaos and uncertainty about what to believe and whom to trust. VOA, with its commitment to fact-based journalism, connections to far-flung audiences, and its vast net¬ work of affiliates, is an effective communications tool against such threats because it can be trusted; trust¬ ed to provide comprehensive news, information and context. Its mission was codified into law in 1976 in the VOA Charter which states that VOA will be a con¬ sistently reliable and authoritative source of accu¬ rate news; that it will tell America’s story, not any sin¬ gle segment of American society, but a balanced and comprehensive projection of significant American thought; and that it will present U.S. policies clear¬ ly and effectively, along with responsible discussion and opinions of those policies. Some argue that in today’s environment that is not enough, that given the efforts of networks such as Russia Today (RT) and China Global Television Net¬ work (CGTN), we need to do more, fight fire with fire. However, truth will always be much more power¬ ful than propaganda in the long run — and the effort to counter disinformation must be just that: a long¬ term play. During World War II, U.S. sailors may have listened to Japanese propaganda disseminated by “Tokyo Rose.” They enjoyed the music, but they didn’t believe what was said because it was so obviously false. I was in China when the post-Cultural Revolu¬ tion control of information was near-absolute. Even then, people understood they did not have the whole truth. Today, Russia’s Sputnik TV is so biased in favor of the Kremlin that its influence on rational observers is muted. Despite the billions spent on their efforts, RT and China’s CGTN have relatively small audiences. There are also indications that falsified, or overly hyped and biased views, can backfire. In Germany, far-right groups have revived the Nazi-era term “Lu- genpresse” or “lying press” to describe the media, in particular, its coverage of refugees and the German government’s immigration policies. However, a new annual survey conducted by the University of Wurz¬ burg shows that German trust in the media actual¬ ly went up sharply in 2016. The level of trust, at 55.7 percent, is the highest since the survey was started in 2000. Even the number of respondents who con¬ sidered themselves right-wing and said they trust the press increased by 18 percent over the last year, rising from 33 percent in 2015 to 51 percent in 2016. Heavily biased content is seen for what it is. The need for accurate, comprehensive, news and information is evident by the fact that many of those in the VOA audience risk their lives just to access it. In China and Tibet, efforts to circumvent government censors are widespread. In Iran, satellite dishes are illegal, but you can find them everywhere. People try to disguise and hide them in order to access VOA and other western content. 62 ACPD I Can Public Diplomacy Survive the Internet? In addition, journalists risk their lives to work for VOA. One reporter who worked for VOA in the Feder¬ ally Administered Tribal Areas (FATA) region and Pa- khtunkhwa province of Pakistan had his house blown up and was eventually forced to flee to the United States. Another was gunned down in a mosque near Charsadda, a town close to FATA. Yet another report¬ er working for VOA in Syria recently had his house bombed by ISIS. Others faced torture and abuse just to have the opportunity to exercise the right of a free press. In parts of the world where alternate sources of in¬ formation are very limited or virtually non-existent, or where systems are highly polarized, VOA is a bea¬ con of light. By providing unfettered access to news and information in 47 languages, VOA reaches peo¬ ple in a way that CNN cannot. In fact, VOA reaches more than three times the domestic (U.S.) audience of CNN, MSNBC and Fox News combined. By the end of 2016, VOA’s global weekly audience stood at 236.6 By adhering to the principles of good journalism, VOA represents significant American thought in¬ cluding demonstrating the value of the rights to free speech and a free press. It also establishes a bulwark against disinformation. VOA is a voice the audience can trust — in good times and in bad. Its authoritative, engaging rapport builds a solid foundation for posi¬ tive, long-term relationships. It takes years to estab¬ lish the high level of trust VOA has earned; but if you take away that credibility and authenticity, trust can disappear immediately. Edward R. Murrow, then Director of the U.S. Informa¬ tion Agency, stated in 1963 in congressional testimo¬ ny, "American traditions and the American ethic require us to be truthful ', but the most important reason is that truth is the best propaganda and lies are the worst. To be persuasive, we must be believable; to be believable, we must be credible; to be credible we must be truthful. It is as simple as that." “VOA reaches more than three times the do¬ mestic ( U. S.) audience of CNN, MSNBC and Fox Newscombined. Bytheendof2016, VOA’sglob- al weekly audience stood at 236.6 mil I ion...” million — the highest number ever recorded and a 26 percent increase over the previous year. The size of the audience, while impressive, is just one part of the story of VOA impact. Research shows that 86 percent of the audience finds VOA to be trust¬ worthy, and three-fourths say VOA helps them to un¬ derstand U.S. policy and current events. The highest-ranking diplomat ever to defect from North Korea, Thae Young-ho, recently said that while still a foreign ministry official in Pyongyang, he read what he termed “reference radio materials of VOA” every morning and afternoon. He said the “North Ko¬ rean regime also pays great attention on the contents of VOA, so I think it is very important that VOA should further strengthen its activity, and also its contents so that, one day, I hope VOA is remembered by North Korean people as a kind of, you know, the main player who contributed a lot for the reunification of the Ko¬ rean peninsula.” From its beginnings, this is what set VOA apart. Whether it be German propaganda in World War II or Soviet propaganda in the Cold War, facts and bal¬ anced reporting were powerful weapons. Fast for¬ ward to state-sponsored false narratives, radical Is¬ lamist propaganda, and a bewildering array of things passing themselves off as fact online. We may not be able to counter every falsehood or half-truth point- by-point, but we can paint a different narrative, one that is truthful and constructive. Compared to the plethora of anti-U.S. propaganda splattered across the internet, VOA provides an alternative canvas of news and information that is believable because it is based on facts. True stories about Americans, dias¬ pora communities, U.S. government and public poli¬ cies, health and technology issues. When VOA tells these stories, the audience often perceives more than we may even realize. Last year, VOA’s Khmer service reporter was providing cover¬ age of pro- and anti-gun protestors at the Republican ACPD I VOA: A Weapon of Truth in the War of Words 63 National Convention via Facebook Live. The crowds were particularly rowdy and tense. However, what really stood out to Cambodian viewers was some¬ thing else. Among their comments on Facebook in real time were “Police in the U.S. do not beat up pro¬ testors like Hun Sen’s police in Cambodia” and “If it was in Cambodia, Hun Sen would send police and dogs to beatand bite the protestors already.” This is just one example of how, by sharing truthful facts, VOA not only explains the context of a news sto¬ ry, but also shows what life in the United States is like. VOA journalists often go to great lengths and face significant risks to get the facts straight. It is the trust in VOA reporters that led the Somali President in Mogadishu and Somali immigrants in the United States to agree to take part in a joint town hall. The same trust and credibility prompted one of Ukraine’s leading television networks to ask a VOA anchor to moderate a parliamentary debate. Whether they are covering wars, natural or man¬ made disasters, telling American stories or explain¬ ing U.S. government policies — VOA reporters and programmers are trusted and reliable. Truth and fact-based journalism are what the audience needs atthistime — andarethestrongestweaponswehave in the war of words. U.S. 201 6 ELECTIONS: A CASE STUDY IN “INOCULATING” PUBLIC OPINION AGAINST DISINFORMATION By Jonathan Henick, Principal Deputy Coordinator for International Information Programs and Ryan Walsh, Senior Advisorfor Digital Product, Bureau of International Information Programs Following the 2016 U.S. presidential election cycle, politicians and pundits sounded the alarms over “fake news” and its potential role in influencing public opin¬ ion. Facebook and Google scrambled to build tools and partnerships to address the more egregious cas¬ es and to shore up the confidence of their consumers and advertisers. While “fake news” remains poorly defined and includes some novel efforts to exploit the new media landscape for political and personal profit, one subset has long been a principal concern for public diplomacy practitioners: state-sponsored disinformation. Throughout much of the Cold War, for example, Soviet propaganda engaged prolifically in such efforts in an attempt to undermine Western principles and sustain support behind the Iron Cur¬ tain. The United States and its allies, meanwhile, em¬ ployed their own tools, programs and platforms — in¬ cluding the Voice of America and other broadcasting affiliates — to counter such disinformation. Since the end of the Cold War certain state actors have continued to invest in traditional broadcasting platforms while also developing new programs and techniques to take advantage of the ongoing trans¬ formation in the media landscape — particularly the emergence of social media. These new techniques include the use of coordinated internet “troll” farms, employed to aggressively disseminate disinformation in an effort to sow mistrust and inflame and exploit societal and political tensions through social media.1 The Voice of America and its sister broadcasting or¬ ganizations have also adapted to the digital age and are producing a steady stream of content designed to inform foreign audiences and correct disinforma¬ tion. At the same time, the U.S. Department of State has become the “world’s leading user of eDiploma- cy,” directly engaging on social media platforms with hundreds of accounts managed from Washington or by U.S. embassies and consulates overseas.2 Today, Department of State maintains approximately 750 individual Facebook and Twitter (400 Facebook and 350 Twitter) accounts worldwide, and when it expe¬ riences peak traffic, the Bureau of International In¬ formation Programs’ most popularcontent is shared acrossasmanyas400ofthese properties. Thatsaid, U.S. public diplomacy efforts have struggled to keep pace with the torrent of foreign state-sponsored dis¬ information. U.S. efforts have been constrained by a number of fac¬ tors including resource limitations, the proliferation of self-described digital “news” outlets unconstrained by fact, and the policy clearance process, but perhaps the greatest challenge has to do with how individuals process new information. We would like to believe that veritas omnia vincit (i.e., truth conquers all) and that rational people will be persuaded by factual evi¬ dence. The reality is that philosophers have long ob¬ served, and academic research strongly suggests, that people are highly susceptible to “confirmation bias.” In other words they ignore information that contradicts, and actively seek out information that confirms, their preexisting beliefs.3 In fact, people who are presented with contradictory information or “facts” often become even more dogmatic in defense of their opinions. Other studies have demonstrated a “primacy effect” in which people are most likely to embrace the first piece of information they consume on a particular subject, particularly when introduced to negative information.4 This is especially true if that information is later reinforced with the help of confir¬ mation bias. Another phenomenon known as “source amnesia,” which prevents people from recalling correctly exactly where, when or how information was acquired, compounds the problem.5 As a result, state-sponsored disinformation amplified through a multitude of distribution channels and widely shared by armies of internet “trolls” has been remarkably effective at influencing public opinion, even when the stories are easily debunked by U.S. -supported or in¬ dependent mainstream media sources. The Bureau for International Information Programs (IIP) — oneofthethree bureaus in thepublicdiploma- cy“family”intheU.S.DepartmentofState — has pilot¬ ed a new approach designed to address this partic- 66 ACPD I Can Public Diplomacy Survive the Internet? ular challenge and to generally improve U.S. efforts to engage and inform foreign publics. Executed in the weeks preceding the 2016 U.S. presidential election, the initial pilot aimed to identify and counter specific state-sponsored disinformation that sought to cast doubt on the legitimacy of the U.S. electoral process and, subsequently, U.S. support for human rights and democratic principles abroad. The effort brought together teams of experts from across IIP including native foreign language speakers, content creators, social media experts and data analysts to develop an in cycle targeted content development approach, il¬ lustrated below: Social Listening Evaluation Content Planning Targeted Advertisements First, team members from the office of analytics equipped HP’s in-house language experts with pub- lically available tools to engage in social listening by examining foreign language social media in a num¬ ber of priority regions to identify when and where state-sponsored disinformation about the upcoming U.S. elections were trending. Once stories surfaced through the social monitoring tools, the language and regional experts reviewed the stories for qualita¬ tive nuance to identify the broad disinformation tar¬ geting the U.S. election process. Identified narratives included false accusations that the United States had denied permission for international election observers to access polling stations. Another trend¬ ing thread suggested that ballot boxes in the Unit¬ ed States were susceptible to fraud and vote tallies could be easily subject to political manipulation. Each of these narratives was targeted at foreign publics to undermine the legitimacy of the U.S. elections and discredit ongoing U.S. efforts to promote democrat¬ ic rights and principles abroad. Importantly, these were emerging narratives that had not yet achieved broad distribution in these media markets. Next, HP’s editorial and video units, as well as its speaker and interactive offices, worked quickly to plan content — both by developing new content and modifying existing articles, videos and interactive programs designed to address the negative narra¬ tives identified earlier. This content did not seek to “counter” or directly refute disinformation, but in¬ stead presented factual and engaging narratives clarifying the election process without reference to the disinformation themes. Much of the content was made available on a specially designed U.S. elections web page on HP’s ShareAmerica website. HP’s analytics team designed and executed a paid social media targeted advertising campaign with the goal of “inoculating” broader audiences in targeted countries before they consumed reports containing the disinformation. Finally, IIP evaluated the ongo¬ ing campaign and repeated the cycle, as needed, to identify new disinformation narratives, develop more content and target new audiences for inoculation. “We would like to believe that... truth conquers all and that rational people will be persuaded by fac¬ tual evidence. The reality is that philosophers have long observed, and aca¬ demic research strong¬ ly suggests, that people are highly susceptible to “confirmation bias. ”” ACPD I U.S. 2016 Elections: A Case Study in “Inoculating” Public Opinion Against Disinformation 67 l/S Election* in Russian Media-SG l1 0 ■ S1SU8S O o Viewing 124 mentions for "US elections in Russian media" from 10/01/2016 - 12/01/2016 Filter by All media outlets.. 90 GR + ’ DATE 4 MEDtA OUTLET HEADLINE AUTHOR ARTICLE IMPACT SEO IMPACT TAGS - to aCTK, Ma^nrit no/voTonity n npooncn* MOCTfmio BwOopM kh Qf Q 11/30/16 r_ Bnk*tM;u* M.ip 9 enoco&oa uaHHnynmtMM kb BbiOopnx a * 43 Hoboctm CUJA Low ■M6opM n CU1A npHOKikno ponop/uion KonM‘«teTno hhot TpjimiM* anA/iiOAnTenoM Mkieaui OECE yoaiw'Miu cmm Am op AflBKCftMAP UiAUUIUBB Mm II OAmOM CTpHWB CMpft hh npoioflur tcpMirrankMo mmctmi bmAopon CUJA - n«... ; dokb ■ Mnpo mjOmpbtb/im bmOukmiot r*omm*coa, ■ CUJA Onarooapn aroMy nputnuy nonnn*ui naaOMpawt naa [jf Q 11/39/1G 2. V/glyiid Hoimn rmu nA OBCE yiNA*ii« HecoeepuKHCTBA ha ewOopax b CUJA Medium • Kjmgtmm pncCKitwin o n**flofwix n CUJA (o BMfiopax np«!k<Aa*«Ta CUJA » /oil. ropy r/t« OTcyTcmoiuuiM M»<SntoftjrrB/iM ot Poccmm} Mm paaoe/wm wmjhmb. hto hi acvroa ohm rooraoTcroyioT wpv.im h psmokpbtmhockmm. . •, - npHBoflwr ero cnoaa TACC . Bmbcto c t»m naprmueMTapMM najaan no/iewtiiu otot flymen H& BW&OpM b fi *■ 11/37/16 ^ (2) Blogs... 3a KynwcaMH auopMKUMCKMX uwoopou 12} Authors... • 31 • MononAbHbix* rpjLWAOn. HaCmoflaio/it.. go home! Low HU aulWinu m ClUA nntotmon nmniuiHnA Above: IIP Analytics Tools Depict Spiking Disinformation and Emerging Narratives in November 201 6 The results of this pilot program suggest that this approach has considerable merit and success in countering state-sponsored disinformation. Specifi¬ cally, over a 10-day period leading up to the U.S. elec¬ tions, IIP delivered over 13 million advertisements to foreign audiences in 20 countries searching for information about the U.S. electoral process in their native language. As a result, the IIP election web page, which included 25 unique pieces of content in six languages, attracted over 300,000 unique article views and over 100,000 unique video views. Perhaps more important than just achieving unique views of IIP content are the accompanying high engagement numbers — a strong indicator of interest by the end user. For example, IIP observed a dramatic spike in average session length: 5 minutes 25 seconds for election articles (compared to a normal average 2 minutes 44 seconds) and 1 minute 25 seconds for election videos (compared to a normal average of 21 seconds). 68 ACPD I Can Public Diplomacy Survive the Internet? ■ * 'm QUESTION MORE Hem Aroenca UK Ruaaar puHiui Buain— Sport Op-tdga In vi won Get the new LG G6. Ptut * fr*« 43' LG Smart TV mrften you Mitch to VMtjor) unlimited. Home n l/134 an poMm United States bars Russian monitors from presidential polls - elections commission Q ** EQ 4 Q ES Q The us Has rejoctod a Russian proposal to send diplomats to monitor the vpconvng president ai elections and some states have oven threatened to bring criminal chargos against any that appear at ballot stations, Russian election officials SHARE AMERICA IDUCATION * CIOMA ISSUtS * «OVt>NUCATAMOCmitOOfrv « AAOUTl/J l NCI SH " Q, American election observers go to other countries to help ensure a free and fair vote that follows local ejection laws. But what happens at U.S. erections? Stay Conn£Ct6<J The United States participates in the Organization for Secunty and Coopera ton in Europe (OSCE). That means the U.S. upholds that group's democratic standards and invites othe* OSCE members to observe American elections. OSCE representatives have monitored U.S. pofts since 2004. During the 2016 elections on November 8, there wM be 120 parliamentarians from different rAiaAFr>*c 1 1 rvilllfSA cra’i'vnc in 17 crv*< T>iaw1I lnr< r>f hai/i cave n^f“F Sgn up for Sha*eAmerica updates straight to your nbo* every weekday Above: Examples of Disinformation and Counter-Narrative IIP Elections Content Notably, subsequent analysis of a major state-spon¬ sored disinformation outlet found that IIP content produced to counter false narratives often per¬ formed on par or better than the disinformation, as measured by Facebook’s publically available data on shares and reach. Specifically, individual IIP election stories averaged 145 public shares to an estimat¬ ed potential audience of 5.4 million per story. The state-sponsored disinformation outlet, on the other hand, achieved an average of 218 public shares per story reaching a potential audience of just 2.9 million. Industry standards suggest that the actual reach of content is 1-2 percent of potential reach, meaning ShareAmerica content was likely seen by an organic audience of 128,000 users, over two times the esti¬ mated average size of the organic audience the disin¬ formation was able to reach (58,000+). 6 One reason for the disparity in audience numbers is that the Department of State has a comparative advantage in leveraging its network of hundreds of social media properties. In fact, each IIP story pro¬ duced to dispel election disinformation was posted, on average, to 29+ individual newsfeeds of other De¬ partment of State social media properties. This has exponential implicationson reach. Forexample, HP’s top performing page, the IIP elections English-lan¬ guage homepage, was distributed in the feeds of over 100 individual Department of State properties by so¬ cial media managers in the field, and generated over 400 unique public feed shares to a potential audience of over 1 6 million people. In addition, the paid distribu¬ tion strategy, in English alone, reached an additional 1.6 million unique Facebook users in key markets vul¬ nerable to disinformation. While these numbers re¬ flect just publically available data from Facebook, and don’t account for “private” (and offline) shares, it is encouraging to see the comparatively strong perfor¬ mance of IIP content relative to the disinformation outlet, particularly considering the potential impact of other external factors, such as the possibility that state-sponsored disinformation may be boosted by fake accounts, or be the benefactor of bot networks programmed to share its stories. These social media metrics suggest that not only did the analytics team and language experts cor¬ rectly identify where and when negative narratives about the U.S. election were trending, but the paid distribution strategy worked by effectively placing content relevant to the news cycle in front of target audiences right when they were most likely to be ex¬ posed to disinformation about the elections. Further, and perhaps most importantly, as demonstrated by ACPD I U.S. 2016 Elections: A Case Study in “Inoculating” Public Opinion Against Disinformation 69 ShareAmerica Elections Content 80000 Weeks Before and After Nov. 9 U.S. Presidential Election Above: Unique views of IIP election content peak the week before the 201 6 U.S. Election increased average session length and engagement metrics, IIP election content resonated with target audiences and, correspondingly, the reach of HP’s positive narratives increased. The spike in engage¬ ments, specifically “shares,” are critical to social net¬ work algorithms for priority placement in the news- feeds of others who did not originally engage with the content — those in “secondary networks.” This not only can result in increased reach of the content, but it can also contribute to the perceived credibility of the content itself, due to the fact that a “share” intro¬ duces the content to the target audience’s secondary network via a newsfeed curated by those who have already opted-in to see one another’s status updates. All Indications are that state-sponsored disinforma¬ tion on social media will remain a serious challenge to U.S. public diplomacy efforts moving forward. With metrics suggesting strong performance of this initial “inoculation” effort, IIP will use this campaign as a case study to demonstrate the power and ef¬ fectiveness of integrating data and analytics to drive content production, precisely target audiences and quantitatively measure results as we continue to in¬ stitutionalize these practices throughout the depart¬ ment’s public diplomacy efforts. ENDNOTES 1. Paul, Christopher & Matthews, Miriam, “The Russian “Firehose of Falsehood” Propaganda Mode,” RAND Corporation, 2016. 2. Hanson, Fergus, “Revolution @State:The Spread of Ediplomacy,” Lowy Institute, 2002. 3. Nickerson, Raymond S., “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises,” Review of General Psychology, Vol. 2, No. 2: p. 175-220, 1998. 4. Thorson, Emily A, “Belief echoes: The Persistent Effects of Corrected Misinformation,” Political Communications, Vol 33, p. 460-480, 2015. 5. Schacter, Daniel L., Joanne L. Harbluk, and Donald R. McLachlan, “Retrieval without Recollection: An Experimental Analysis of Source Amnesia,” Journal of Verbal Learning and Verbal Behavior, Vol. 23, 593-61 1 (1984). 6. The Social Media Examiner reports that organic reach is approximately 2 percent of a Facebook page’s fan base for public pages with 500,000+ page likes. (See Smith, Mari. 2017. “How to Maximize Facebook Reach.” <http://www.socialmediaexaminer.com/ how-to-maximize-facebook-reach/>. Accessed 20 April 201 7.) f ' # . Jk s» ♦ * fv \ V INDEFENSE OFTRUTH,ANDTHE THREATOF DISINFORMATION By Jason Stanley, Jacob Urowsky Professor of Philosophy, Yale University Thereisan international, anti-democratic, nationalist movement buoying authoritarians abroad, threaten¬ ing to end hopeful democratic moments in Eastern Europeandelsewhere.Howmuchisduetonewtech- nology delivering novel means of propaganda? And how much of it is the pendulum of history, returning us to age-old concerns about the stability of liberal democratic states? In this essay, I place these con¬ cerns in historical and philosophical context, to eluci¬ date both the problem and the best response. In section I, I begin with the problem of defining the topic. What is propaganda? In section II, I trace the roots of the problem back to the founding texts of western philosophy. Insection III, Igiveadefense of truth, and further expand on the risks of disinforma¬ tion efforts. I conclude with some recommendations about how public diplomacy can be marshaled to re¬ spond to the problem of authoritarian propaganda. SECTION I: PROPAGANDA DEFINED One of the problems with the current debate about “fake news” and propaganda is the lack of a clearthe- oretical taxonomy. I begin this section by explaining the difficulty of characterizing the topic of propagan¬ da. Using definitions from my 2015 book How Propa¬ ganda Works, I attempt what I hope to be a more use¬ ful definition of propaganda. In Section III, I use this definition tocharacterizeauthoritarian propaganda. It might be thought simple and straightforward to characterize our topic, but it is useful to look at some candidate definitions of propaganda to see that it is morecomplexthanonemayinitially realize. First attempt to define propaganda: Propa¬ ganda is the manipulation of public opinion. This is a familiar characterization of propaganda and, yet, it is uninformative. Any attempt to persuade a public of something involves giving an argument of some kind. In defining propaganda, we want to know what the difference is between giving a propagandistic argument, and giving a non-propagandistic argument. This definition tells us that propagandistic arguments are ones that persuade by “manipulation”. But what is it to persuade by manipulation? Here is a natural characterization: Manipulation: Manipulation is the use of devious methods to get an audience to do one’sbidding. What are “devious methods”? One might answer that devious methods are those that involve propaganda. In short, our first attempt is thoroughly uninforma¬ tive. Let’s try again. Second attempt to define propaganda: Propaganda is cherry-picking facts. But what is “cherry-picking facts”? Is “cherry picking facts” not listing all the facts? But it is impossible to list all facts. If someone were to try to list all facts, it would take more seconds than there are in the uni¬ verse. Any list of all facts would have to, for example, list the facts about all the molecules presently found under my left pinky fingernail. Whenever one pro¬ vides information, one selects some facts to present and neglects others. This suggests a second defini¬ tion of “cherry-picking facts.” One cherry-picks facts when one omits relevant facts. But this is both too broad and not broad enough. It is too broad, because it includes cases of propaganda that include just bad arguments (some people are just wrong about what isarelevantfactj.Andevenifweemendthedefinition to “intentionally omits relevant facts,” it omits cases of propaganda that involve false claims, which are of course not facts at all. Here is another definition that overcomes some of these weaknesses: Third attempt to define propaganda: Pro¬ paganda consists of known falsehoods. However, propaganda can be true. Suppose an an¬ ti-Semitic leader gives a talk in front of their parlia¬ ment, bringing out victims of (for example) predatory lending practices by bankers who happen to be Jew¬ ish. Let’s suppose there are victims of such practices by bankers who happen to be Jewish, as is plausi¬ ble — though of course people of Jewish faith are no 72 ACPD I Can Public Diplomacy Survive the Internet? more likely to perpetrate such financial impropriety than people who are not of Jewish faith. And let’s suppose the people being presented as victims are indeed victims of such predatory lending practices. It is still propaganda to present them as such, since it suggests that there is some distinctive problem with the Jewish faith. Propaganda does not need to con¬ sist of known falsehoods. A series of perfectly true assertions can constitute propaganda. Omission of crucial information is char¬ acteristically propagandistic. If I am only told that the blue tribe killed my grandfather, I might form a lifelong hatred of members of the blue tribe. But if I am also given the information that my grandfather enjoyed murdering children from the blue tribe for sport, then I will instead develop a more nuanced view of my personal history. Just as a series of true statements can be clearly pro¬ pagandistic, a series of perfectly false assertions can be clearly non-propagandistic. In teaching physics, one may spend one semester on Newtonian Mechan¬ ics. Newtonian mechanics is false. It is nevertheless a good theory to teach, since it is approximately enough true of middle-sized physical objections. The method¬ ology of the natural sciences is governed by the ideal. It is natural to slip into teaching an ideal model as if it were reality. The divergences between the ideal mod¬ el and reality are not important for the science. Still, if one is teaching an ideal model of physics or chem¬ istry, or even rationality, one is saying false things. Physics classes are not thereby propaganda. Nor is deception necessary for propaganda. It is un¬ questionably true that Hitler was a deeply committed anti-Semite. Nevertheless, in MeinKampf, Hitler is very clear that he is also using anti-Semitism propagan- distically. Is propaganda the use of words to skew debate? It really is not possible to use any word without “skew¬ ing debate” in some way. In his essay, “General Se¬ mantics and Propaganda,” published in 1939, S. I. Hayakawa writes: “In fact, there is nothing that can be named, let alone described, without invoking the wraiths of an entire contextual system. What is ‘money’? What is a ‘house of cor¬ rection’? What is a ‘professor’? What is a ‘musician’? ... a ‘tom-boy’? ... a ‘mortgage’? ... a ‘cat’?”1 Here are some definitions of propaganda from my own work:2 Political propaganda: An argument that employs a political ideal in the service of a goal, seeking to advance or undermine that ideal by non-rational means. The most central kind of political propaganda is what I call undermining propaganda. Undermining propaganda: An argument that employs a political ideal to undermine that very political ideal. Given my definition, propaganda can be either good or bad. It is good when it is used to undermine bad ideals, and bad when it is used to undermine good ideals. Assuming the ideals of liberal democracy are good, we can characterize demagoguery as: Demagoguery: An argument that employs democratic ideals to undermine demo¬ cratic ideals. These are unfamiliar characterizations of propagan¬ da. I have argued that the familiar ones do not help us characterize the terrain. More persuasively, these definitions allow us to see that the structure of dema¬ goguery we face today fits straightforwardly into the model I have outlined. The journalist Peter Pomerant¬ sev characterizes the “political system in miniature” of Vladislav Surkov, the author of Putin’s propaganda regime, as “democratic rhetoric and undemocratic intent.”3 Now that we have defined the terrain, we can begin with an overview of its history. I will argue that the ef¬ ficacy of the propaganda of tyranny is not a byprod¬ uct of novel technologies. It is rather, historically, the chief obstacle to the stability of democracies. What we see in Eastern Europe today, for example, is the fragility of democracy when confronted with some of its chief obstacles. SECTION II: PROPAGANDA, IDEOLOGY AND DEMOCRACY Plato and Aristotle both regarded stability as a vital metric by which to evaluate political systems, though they differed on their judgments about democracy. Plato’s Republic is about proper governance of “the city” and “the soul,” and includes a description of “the ACPD I In Defense of Truth, and the Threat of Disinformation 73 characteristics of democracy,” such as “the city’s tol¬ erance.” In summary, “it would seem to be a pleasant constitution, which lacks rulers and not variety and which distributes asort of equality to both equals and unequals alike.”4 A culture whose central value is liberty will lead to sweeping social equality. In a democratic city, stu¬ dents in the academies challenge their teachers. A democratic culture equalizes those who are natu¬ ral-born and immigrant; in such a system “[a] res¬ ident alien or a foreign visitor is made equal to a citizen.” Democracy is inconsistent with enslaving others. And in a democracy, there is equality between men and women.5 Socrates recognizes that the flourishing of liberties, the diversity of practices and customs, and social equality, may seem attractive. However, he urges us to attend to its risks. People are not naturally inclined to self-governance, “always in the habit of setting up one man as their special champion, nurturing him and making him great.”6 Democracy also creates a The equal participation of all citizens in the formation of the policies that will be adopted and fairly applied lends the system its stability. Aristotle also emphasiz¬ es democracy’s epistemic virtues, arguing that open and honest cooperative deliberation about policy be¬ tween all citizens yields better results, in the form of wiser policy, further strengthening the stability of the system. Democracy requires a clean public square. Plato’s democratic city is based upon a notion of liber¬ ty as unconstrained freedom to satisfy one’s desires, freedom from the limitations of customs and tradi¬ tions. Aristotle’s conception of democracy, by con¬ trast, allows democratic societies to have communal values. However, this is possible only if all citizens freely and equally participate in the decision to adopt them, decisions that must be continually revisited. Participating equally in such decisions is, for Aristo¬ tle, genuine freedom. Contemporary liberal democracies differ from these conceptions of democracy in at least two ways. First, they incorporate essential insights of Christianity, “Plato sees in democracy’s ideal of the free¬ dom of speech the cause of its potential downfall. ” vast amount of resentment, due to the social upheav¬ al required by prizing freedom and the attendant costs to traditions, customs and hierarchies. Plato sees in democracy’s ideal of the freedom of speech the cause of its potential downfall. Pressure for freedom and equality leads to resentments of fel¬ low citizens, as will the inevitable hypocritical use of these ideals (e.g., when the ideal of liberty is used to justify corruption). These resentments can be exploit¬ ed by outside forces to stoke fear of fellow citizens. Since tyranny is liberal democracy’s greatest enemy, the propaganda of tyranny characteristically takes the form of undermining propaganda, and what is of¬ ten referred to these days as disinformation. Aristotle was more sanguine. In Aristotle’s demo¬ cratic city, all citizens participate in the formation of the laws by which they are governed, an activity that for Aristotle was the purest expression of freedom. such as the concept of human rights. Secondly, they involve elected representatives to act on behalf of our best interests, tasked to deliberate with one another reflectively, openly and truthfully, with willingness to changing their minds and compromise. American democracy differs in a significant way from most other Western democracies, which make Plato’s concerns particularly relevant. Democracies throughout the world, in the words of Jeremy Wal¬ dron, have the “conviction that a liberal democracy must take affirmative responsibility for protecting the atmosphere of mutual respect for its citizens.” But our Constitution provides the broadest protec¬ tions for speech in the political arena. India’s first amendment bans hate speech; our first amendment protects it. If Plato is right, with the rise of ubiquitous, mobile connectivity, the global public sphere is at risk of being overrun with competing resentments from around the world. 74 ACPD I Can Public Diplomacy Survive the Internet? There is much attention that has been given to the force of technology; and it is true that we have seen a new way to target specific voters by ideology by observing their online habits. But we also face an old problem in new form. We speak now about how the internet has unleashed the tide of free expression, bringing with it supposedly novel dangers. Yet Victor Klemperer, in his 1957 book The Language of the Third Re¬ ich, writes, about the Weimar Republic: “The Republic, almost suicidally, lifted all controls on freedom of expression; the National Socialists used to claim scornfully that they were only taking advantage of the rights granted to them by the constitution when in their books and newspapers they mercilessly attacked the state and all its in¬ stitutions and guiding principles using ev¬ ery available weapon of satire and belliger¬ ent sermonizing. There were no restraints whatsoever in the realm of the arts and sci¬ ences, aesthetics and philosophy. Nobody was bound to a particular dogma or ideal of beauty, everyone was free to choose. This motley intellectual freedom was cele¬ brated as a tremendous and decisive leap forward compared with the imperial age.”7 As we have seen, Plato is clear-eyed about the risks that certain forms of propaganda pose to liberal democracy; free expression allows for the airing of views that inflame and divide the public against one another, leading to tyranny. Nor was this point un¬ known to democracy’s greatest enemies in the 20th century. Joseph Goebbels said freedom of expres¬ sion “will always remain one of the best jokes of de¬ mocracy, that it gave its deadly enemies the means by which it was destroyed.”8 The contemporary democratic system of the West is hardly the first to face challenges posed by its most cherished values, nor will it be the last. The virtues of democracy — the ever-expanding circle of liberty, encompassing women, religious minorities, gays and other groups — are evident to many. But philosophers from Plato through Hobbes and Rousseau have ar¬ gued that its commitment to liberty is likely to render it less stable than authoritarian systems. Yet, this very weakness is also its greatest strength. Aristotle, in the Politics, paves the way for democrat¬ ic stability, arguing that a genuine commitment to equality makes society less susceptible to revolution. Aristotle’s conception of equality is political equality. Aristotle argues that a society in which each citizen is fully represented in public debate will not lead to a breeding ground for anti-democratic resentment. If Aristotle is right, the greatest advertisement for our democratic system abroad is afull and open commit¬ ment todemocratic participation by all of our citizens. Toadvertise democracy is to advertise a system with easy access to the ballot box, where public disputes are aired openly in an atmosphere of transparency. Hypocritical employment of our values will be seized uponbyouradversariesasevidencethatdemocratic valuesareonlyevermasksforinjustice. SECTION III: IN DEFENSE OF TRUTH The eminent ethicist Stephen Darwall describes a well-constituted democratic society as one “in which people are answerable to one another for their con¬ duct ... one that values public inquiry, getting at the truth behind social appearances and ‘speaking truth to power’ . . . When we . . . respect all equally ... we com¬ mit ourselves to a mutual accountability that implicit¬ ly honors fact over appearance.”9 Truth underlies the democratic ideal of equal re¬ spect. Without truth, there is no way to speak truth to power. Truth underlies dissent. Without truth, there is no way to dissent by appealing to facts that under¬ mine the authority of a leader. T ruth underlies trust. Without trust, our institutions cannot function; their authority merely will rest on power. That is not dem¬ ocratic authority. Democratic and cooperative systems depend on truth, because truth underlies equal political equality. Truth and falsity, indeed reality, are the referees in the public arena. If the public arena is guided by truth, someone lacking material power can nevertheless be a political equal, since they can appeal to facts against those with more material power. Truth is the essential backbone of a democratic society. Given the foregoing, what, then, is the shape and form of disinformation? Characteristically, disinformation takes the form of the undermining of reality. Exam¬ ples in include efforts to weaken public confidence in democratic institutions, or the establishment news sites using false information to undermine legitimate news institutions. Media outlets masquerading as news, like the numerous portals that popped up in Macedonia in 2016, are perfect examples of this. They appeal to the ideal of objective truth to undermine objective truth. ACPD I In Defense of Truth, and the Threat of Disinformation 75 CONCLUSION Public diplomacy practitioners have their work cut out for them, but should also feel reassured in that the challenges we face today are, to a large extent, rooted in the deep history of human civilization. Dem¬ ocratic systems, by the very nature of being open societies, are more vulnerable to foreign efforts to spread disinformation. Modern technologiesamplify this threat, at least at first. Insofar as public diplomacy aims to support dem¬ ocratic systems of governance abroad, emphasis must be placed on restoring confidence in the value of widespread public participation in politics, and a faith in transparent intuitions to be capable and/ or redeemable in serving the public. In short, public diplomacy needs to confront the cynicism that is, in part, driven by the modern media ecosystem (the or¬ igins of which are detailed in Sam Ford’s essay in this report). Encouraging civil, respectful discursive engagement needs to also be a priority, given the threat of uncivil discourse presents to democratic systems. Empha¬ sizing this point in exchange program curriculum should be prioritized, for example. Embassies and in¬ ternational broadcasters can both contribute to this through their public programing around the world. Models of democratic debates on important public policy issues that embody respectful, fact-based back and forth and consensus building are a good way to remind foreign citizens, and ourselves, that discur¬ sive disagreements need not aggravate social cleav¬ ages. ENDNOTES 1. S.l. Hayakawa, “General Semantics and Propaganda.” The Public Opinion Quarterly 3, no. 2 (1939): 197-208. 2. Jason Stanley, How Propaganda Works (Princeton: Princeton University Press, 2015). 3. Peter Pomerantsev, “The Hidden Author of Putinism: How Vladislav Surkov invented the new Russia”, The Atlantic, November 7, 2014. 4. Plato’s Republic, p. 558b/c. 5. Plato’s Republic, p. 562e/563b. 6. Plato’s Republic, p. 565d. 7. Victor Klemperer, The Language of the Third Reich (Max NiemeyerVerlag, Halle, 1957). 8. Paul Jospeh Goebbels was a German politician and Reich Minister of Propaganda in Nazi Germany from 1 933 to 1 945. Goebbels is quoted in Karl Dietrich Bracher, Manfred Funke, and Hans-Adolf Jacobsen, Nationalsozialistishe Diktatur, 1933-1945: Eine Bilanz (Bonn: BundeszantralefurPolitischeBilding, 1983),p. 16. 9. Stephen Darwall, Honor, History, and Relationship: Essays in Second-Personal Ethics II (Oxford, Oxford University Press, 2013). 'll' ^|||P 'll' III 'll' III UNITED STATES INFORMATION AGENCY JW7 aI J) 4 % w. ^ / \ I ft L/o Max, Peter, 1937, from the Library of Congress PUBLIC DIPLOMACY AND STRATEGIC NARRATIVES1 By Laura J. Roselle, Professor of Political Science and International Studies, Elon University This essay makes five interrelated arguments about the efficacy of public diplomacy efforts: (1) public di¬ plomacy cannot be understood without understand¬ ing the importance of narratives in social and polit¬ ical relations; (2) public diplomacy involves shared narratives created with publics abroad; (3) a new communication ecology undergirds public diploma¬ cy efforts; (4) the “post-truth society” is a narrative that focuses attention away from public diplomacy; and (5) multi-method analyses are needed to under¬ stand narrative creation, diffusion and effects. PUBLIC DIPLOMACY CANNOT BE UNDERSTOOD WITHOUT UNDER¬ STANDING THE IMPORTANCE OF NAR¬ RATIVES IN SOCIAL AND POLITICAL RELATIONS. A narrative is “a sequence of events tied together by a plot line” and is a social product produced within a social context.2 Narratives are central to the way hu¬ man beings think. They are important to people as conceptual organizing tools that allow individuals to understand one another within a particular context. The importance of narratives is recognized by nu¬ merous fields including political science, psychology, anthropology and sociology. Strategic narratives are defined as “a means for po¬ litical actors to construct a shared meaning of the past, present and future of politics in order to shape the behavior of other actors.”3 Debates over the environment, energy provision, re¬ form of global institutions, security and power tran¬ sition can all be understood through the lens of stra¬ tegic narrative. Each proposal to confront problems of the international community is driven by underly¬ ing narratives that may be strategically deployed by actors. This is a complex endeavor as the world is marked by contestation over narratives, but a com¬ pelling narrative may become a power resource on its own. There are three different types of strategic narra¬ tives that we identify: • International system narratives describe how international order is structured, who the players are and how the system works. For example, a Cold War narrative suggests a bi-polar international order marked by con¬ flict between two competing powers (and most often even now these countries are identified as the United States and the Rus¬ sian Federation). • Identity narratives describes the political ac¬ tor, what values it has and what goals it has. Narratives about what led to the creation of NATO, its values and what goals it has today, would be an example. • Policy narratives set out why a policy is need¬ ed and how it will be implemented. This in¬ cludes narratives that seek to persuade peo¬ ple to support a particular policy or action. Usually policy narratives reference, at least implicitly, identity and system narratives to set the policy within a context to enhance its legitimacy. It is important to recognize that these different types of strategic narratives can complement or under¬ mine each other. If a policy, for example, does not seem to be in accord with a state’s identity narratives, support for that policy may suffer as a result. In addi¬ tion, actions taken by a state can undermine broad¬ er narratives that are meant to support longer-term goals about constructing a shared understanding of how the international system should function. For ex¬ ample, a post-Cold War system narrative that might have incorporated a greater reliance on cooperation and diplomacy was undermined by U.S. and Russian military actions in Iraq and Chechnya, respectively.4 PUBLIC DIPLOMACY AS SHARED NARRATIVES A dominant strain of academic literature makes some crucial points about public diplomacy. First, public di¬ plomacy is designed to “foster mutual trust and pro¬ ductive relationships,” typically for a strategic pur¬ pose.5 This implies that the goal of public diplomacy is the enhancement of soft power. Rather than focusing on hard power as the ability to coerce or induce an¬ other to do something, scholars and politicians often 78 ACPD I Can Public Diplomacy Survive the Internet? say that soft power is the ability to influence others through the attraction of culture, values, narratives and policies — which are soft power resources.6 A dif¬ ferent way to think about soft power is as the ability to create consensus around shared meaning. Creat¬ ing a shared consensus, however, can be much more difficult than using hard power to force another to do something, but there is reason to believe that the re¬ sults can be more lasting. Soft power resources may set the stage for shared understandings and this en- timing and audience as political actors (including in¬ dividuals), non-state actors, NGOs, terrorist cells and international organizations have access to communi¬ cation technologies that will reach a vast audience.8 Soft power may be a resource on which leaders can draw; however, skilled political leadership is still re¬ quired as soft power is employed in foreign policy and international relations. As Richard Holbrooke once commented to Michael Ignatieff, “Diplomacy is not like chess. . . It’s more like jazz — a constant improvisa- “Creating a shared consensus, however, can be much more difficult than using hard power to force another to do something, but there is reason to be¬ lieve that the results can be more lasting. ” hances other types of interactions, including oppor¬ tunities in enterprise and coordination of shared hu¬ man goals, such as the alleviation of human suffering. Second, public diplomacy implies listening and cre¬ ating narratives with foreign publics. This suggests that the strategy of narrative construction should be collaborative. Public diplomacy does not imply that a narrative is constructed in isolation for political rea¬ sons to be injected into a foreign population. As act¬ ing under secretary for public diplomacy and public affairs, Ambassador Bruce Warton notes: “Crafting and effectively putting forth that narrative with for¬ eign publics is the real challenge of public diplomacy today.”7 How are narratives constructed with foreign pub¬ lics? That is the crucial question that should guide thinking on public diplomacy. A NEW COMMUNICATION ECOLOGY UNDERGIRDS SOFT POWER POSSIBIL¬ ITIES The importance of understanding strategic narra¬ tives as being mutually constructed is even more important in our new communication ecology. New means of communication and the greater ability of people around the world to access these new com¬ munication technologies shape public diplomacy to¬ day. Elites have lost relative power over information, tion on a theme.” The ability to devise and implement a coherent strategy rests on the vagaries of events and the views of others.9 It also rests on collaboration rather than unilateral, one-way communication. “POST-TRUTH SOCIETY” AS A STRATE¬ GIC NARRATIVE Yet, a review of the popular media and some scholar¬ ship done on new communication technologies, sug¬ gests the idea — the narrative — that we live in a “post¬ truth society.”10 This narrative, by asserting that there is no desire for, or focus on, “truth,” actually under¬ mines the ability to construct strategic narratives based on shared understandings of international order and policy. The post-truth narrative is, itself, both strategic and problematic from the perspective of conducting effective and ethical public diplomacy. Instead of a post-truth society, it may be more apt to say that we are living in a world with more narrative contestation. There are more ways to construct and share narratives. This does not necessarily mean that the truth is less important, but this does point to the need to recognize that public diplomacy will succeed only when multiple narratives are recognized and un¬ derstood. This suggests that it is important to under¬ stand how different people and groups experience the world in different ways, and that the acknowl¬ edgement of multiple perspectives and narratives can undergird the legitimacy of public diplomacy. We should note that there are still facts in the world, and ACPD I Public Diplomacy and Strategic Narratives 79 lies or falsehoods can and should be challenged, but multiple perspectives and narratives may be helpful in understanding an issue or policy. It may also mean that people in democracies need to become more skilled in reading power in commu¬ nication messages. Experience in the Soviet case is informative. In the Soviet Union people were exposed to the repetition of media messages controlled by the government, but we know that the hypodermic mod¬ el of media effects did not work in this context.11 Sovi¬ et citizens learned to “read” the media messages not as presenting the “truth,” but as presenting the sto¬ ry that the powerful wanted to project. In the Soviet case, overbearing, controlled messages did not work in the long run. People learn to read power relations. That citizens in the West might not be good at this is interesting, but hardly surprising — and it may be changing. Evidence from Q-sorts,12 done in issue areas as diverse as the environment and U.S./U.K. policies discussions about Syria, show that people do not swallow narratives whole, but craft their own narratives from the broadcloth of existing narratives andfromtheirownlivedexperiences. MULTI-METHOD ANALYSES ARE NEED¬ ED TO UNDERSTAND SOFT POWER There are many new ways of monitoring, measuring and evaluating the impact of strategic narratives in a new media environment. Examples of quantita¬ tive measures include: analyses of reach, time spent with online content, number of Twitter followers and retweets and positivity of sentiment. However, these may not capture the quality of engagement and what follows from it. Additionally, these may not capture the patterns of public narratives that do not match elite narratives. If public diplomacy is to be most effective in increasing U.S. attractiveness, strategic narratives must be constructed with an understanding of sys¬ tem, identity and policy narratives within the public realm. For example, a policy narrative is usually tied to system and identity narratives that seek to place the policy within a specific context. Additional qual¬ itative research is needed — including focus groups, interviews and participant observation. A multifac¬ eted approach is needed to fully understand the use and effectiveness of public diplomacy. For example, using Q-sort methodology can show how policy narratives may differ between elite and the public. For example, in an analysis of U.S. and U.K. narratives about potential policies towards Syria in 2013-2014, six elite narratives were found in both cases, but six differently constructed U.K. respon¬ dents’ narratives and four differently constructed U.S. respondents’ narratives were found.13 That is, there were patterns to respondent narratives but they did not replicate or match elite narratives. In the U.K. respondents’ narratives there was agreement in a number of narratives that the international com¬ munity has a responsibility to uphold international law (system), but whether or what action to take was unclear (policy), and British leadership (identity) was supported strongly in only one narrative. In the U.K. case, most respondent narratives opposed interven¬ tion saying it might do more harm than good — which was shared in many of the elite narratives found in the House of Commons debate — but there was dis¬ agreement about what that might mean for the U.K. and the world more broadly (identity and system nar¬ ratives). So, it is important to note that there may be different narratives even among those who may support a spe¬ cific policy. In Q-methodology consensus statements are those that are shared between different narra¬ tives. We found no consensus statements among U.K. respondents’ narratives in the 2013-2014 study. This confirms the picture of U.K. public attitudes to for¬ eign affairs being particularly ambivalent during this period of time. In the U.S. case, there was consensus around six basic statements about foreign affairs within the four U.S. respondents’ narratives. For ex¬ ample, all four narratives shared disagreement with the statement that U.S. intervention in Syria would radicalize American Muslims and bring terrorism to the streets of the United States. The consensus statements give a perspective on shared component parts of what may be different narratives. Looking for similarities in system and identity narratives among those who disagree on policy, and looking for differ¬ ences in system and identity narratives among those who agree on policy, offer insight into future support and legitimacy for policy proposals. In the U.S. case, no respondents’ narrative support¬ ed the statement that the United States must act as leader of the international community, although one supported the statement that the world is looking to the United States for action. One narrative strongly supported the idea that the 2003 Iraq war showed the difficulty of using military intervention, while all narratives agree, to varying degrees, that interven¬ tion could lead to escalation. There was strong sup¬ port for humanitarian concerns in one narrative. Two 80 ACPD I Can Public Diplomacy Survive the Internet? narratives in the study agreed that domestic politics was driving U.S. policy, but these narratives did not include support for unilateral congressional action or support for calling President Obama an imperial president. Knowing something about identity narra¬ tives (what the United States is and should be) and system narratives (the structure of the international system itself) allows a more refined understanding of support for specific policies. It is important to understand how public narratives change over time. It is also important to recognize that there are multiple (but not an unlimited num¬ ber of) public narratives, and that they may or may not mirror elite narratives. Understanding system and identity narratives would allow those involved in public diplomacy to construct policy narratives that speak to, or at least acknowledge, those underlying system and identity narratives. The Russian Achilles’ heel is that the Russian system and identity narra¬ tives are exclusive, shutting others out by asserting that those with power should do what they want. U.S. public diplomacy — conceived as listening and collab¬ orative — avoids that Russian vulnerability. Overall, public diplomacy should focus on under¬ standing system, identity and policy narratives with audiences around the world. In addition, audiences will be more discerning about media messages when they appreciate the power relations involved in media messages. Finally, multi-method analyses are needed that focus on audience narrative con¬ struction in a nuanced and subtle way. ENDNOTES 1. Part of this text is derived from Laura Roselle, “Written Testimony to the U.K. House of Lord’s Select Committee on Soft Power and U.K. Influence.” 2. Cristina Archetti, “Narrative Wars: Understanding Terrorism in the Era of Global Interconnectedness.” In Forging the World: Strategic Narratives and International Relations. University of Michigan Press, 201 7, 21 8-245, 220-221 . 3. Alister Miskimmon, Ben O’Loughlin, and Laura Roselle, “Forging the World: Strategic Narratives and International Relations.” University of Michigan Press, 2017. 4. Laura Roselle, “Strategic Narratives and Great Power Identity,” in Alister Miskimmon, Ben O’Loughlin and Laura Roselle, eds. Forging the World: Strategic Narratives and International Relations. (Ann Arbor: University of Michi¬ gan, 2017, 56-84). 5. https://uscpublicdiplomacy.org/page/what-pd 6. Nye sets out culture, values, and policies as important resources of soft power, http://www.foreignpolicy.com/ar- ticles/2006/02/22/think_again_soft_power Of course, these are not distinct, as all are mutually constructed. I add narratives to the list of soft power resources. Conceptually “narratives” focus attention on communicative process¬ es associated with soft power. 7. https://www.state.gOv/r/remarks/2Q1 7Z268592.htm , emphasis added. 8. Robin Brown, “Getting to War: Communication and Mobilization in the 2002-03 Iraq Crisis,” in Media and Conflict in the Twenty-First Century, edited by Philip Seib (New York: Palgrave, 2005); Andrew Chadwick, The Hybrid Media System: Politics and Power (Oxford: Oxford University Press, 2013); Monroe E. Price, Media and Sovereignty: The Global Information Revolution and Its Challenge to State Power (Cambridge: MIT Press, 2002) 9. Miskimmon, O’Loughlin, and Roselle, 2013, 69. 10. For example, see: Francis Fukuyama, “The Emergence of a Post-Fact World,” Project Syndicate, Jan. 12, 2017, https:// www.project-syndicate.org/onpoint/the-emergence-of-a-post-fact-world-by-francis-fukuyama-2017-01; Peter Pomerantsev, “Why We’re Post-Fact,” Granta, July 20, 2016, https ://g ra nta . co m/why-were- post-fact/; Kathleen Hig¬ gins, “Post-Truth: A Guide for the Perplexed,” Scientific American, Dec. 5, 2016, https://www.scientificamerican.com/ article/post-truth-a-guide-for-the-perplexed/ ACPD I Public Diplomacy and Strategic Narratives 11. See the work of Ellen Mickiewicz on this point. 12. Q-methodology allows the study of this remixing of narratives. The aim is not to understand how individuals process information per se, but rather to see how narratives morph and change in the midst of contestation and competition to form new common narratives. Respondents are asked to assess various component parts of narratives (Q-sort) and factor-analysis is used to look for patterns in how new/different narratives are constructed. Ben O’Loughlin, Laura Roselle, and Alister Miskimmon. “Public Narratives about Syria: A Q-Sort Analysis in the U.K. and U.S.” Paper presented atthe ISA Annual Conference, New Orleans, 1 8-21 February 201 5. 13. The Q-sort sought to understand youth narratives about Syria policy in each country. Ben O’Loughlin, Laura Roselle, and Alister Miskimmon. “Public Narratives about Syria: A Q-Sort Analysis in the U.K. and U.S.” Paper presented atthe ISAAnnual Conference, NewOrleans, 1 8-21 February201 5. CRAFTING RESILIENTSTATENARRATIVESINPOST TRUTH ENVIRONMENTS: UKRAINE AND GEORGIA By Vivian S. Walker, Professor of National Security Strategy, National War College The rapid evolution of communications paradigms, as well as vulnerabilities created by unlimited and un¬ filtered access to information, challenge a state’s abil¬ ity to craft a credible narrative about its interests and aspirations in the service of its strategic goals. 1 First, difficulty in discerning objective fact from subjective belief in a “post-truth” information environment de¬ grades narrative authenticity.2 Moreover, the erosion of public trust in state institutions and traditional media sources further damages a state’s capacity to make its case in the public sphere. Russia has taken advantage of this overloaded and compromised information space to launch punitive disinformation campaigns against former satellite states seeking lasting relationships with Euro-At¬ lantic institutions. Russian information attacks force these vulnerable emerging democracies to confront existential questions about national identity, values and models of governance. To neutralize the toxic and often destructive effects of Russian propaganda, targeted countries must project a coherent, consis¬ tent account of their unique political, economic and security assets. This counter narrative must also es¬ tablish the state as a resilient security and economic power in the region. A comparison of Russian disinformation effects in Georgia and Ukraine offers useful insights into the challenges associated with the creation of viable state narratives in a post-truth environment. Russia’s weaponization of information has recently attracted agreatdeal of international scrutiny, especially in the aftermath of Putin’s triumphal annexation of Crimea and the occupation of two Eastern Ukrainian prov¬ inces.3 Less well documented, but equally troubling, is the ongoing information war being waged in Geor¬ gia. Russia’s 2008 invasion, in which it took control of two Georgian territories, provides a chilling coun¬ terpoint to its powerful, and potentially destabilizing, disinformation campaign to bring Georgia back into itssphereof influence. Historically Georgia and Ukraine have been at the mercy of aggressive regional powers and competing religious and cultural influences. For centuries, both countries experienced brief periods of sovereignty interspersed with long stretches of conflict. Russia and Turkey, for example, have treated Georgia as a pawn in a series of attempts to assert regional domi¬ nance, much as Ukraine has been subject to a series of invasions and occupations by Poland, the Crimean Khanate, Hapsburg Austria and Tsarist Russia. Both countries enjoyed a taste of independence before be¬ ing swallowed into the Soviet Union in the aftermath of the Russian revolution. Following the collapse of the USSR, Georgia and Ukraine began the slow process of democratic insti¬ tution building and political integration with Western institutions. Georgia’s 2003 Rose Revolution, which resulted in a peaceful transition of power, was her¬ alded as a “new wave of democratization” for the region.4 Similarly, the 2004 Orange Revolution in Ukraine upheld and validated the power of the elec¬ toral process and civil resistance. But Russia’s 2008 invasion of Georgia and subsequent occupation of South Ossetia and Abkhazia, followed by the 2014 an¬ nexation of Crimea, highlighted the fragility of these gains. Today Georgia, like Ukraine, is a country under actual Russian occupation. Their sovereignty has been compromised, and the threat of sustained or renewed conflict with Russia has limited and simulta¬ neously polarized, their foreign policy options. At the same time, both countries remain at war in the infor¬ mation space, vulnerable to Russia’s adroit manipula¬ tion of facts and ability to exploit audience paranoia and predilections. RUSSIAN DISINFORMATION EFFECTS: IDENTITY Into that space between East and West, between tra¬ ditional and modern cultures, between illiberal and liberal political institutions, Russia inserts an insidi¬ ous and potentially undermining series of messages about the supremacy of the “Russian World.” These include a call for a return to the mythologized ver¬ sion of a “Greater Russia;” a reminder of the target country’s place historical and cultural place in the Russian world; the promotion of Euro-skepticism along with an effort to discredit the target country’s European (EU) and Euro-Atlantic (NATO) aspirations; 84 ACPD I Can Public Diplomacy Survive the Internet? a rant against an aggressive and hostile “West” that threatens Russian’s regional security and economic interests; a tendency to blame current conflicts and global economic threats on selfish Western nations corrupted by their national interests; and an appeal to a pan-Slavic orthodoxy as an antidote to corrupt and overbearing Western values. 5 As part of its anti-Western discourse, Russia’s broad information warfare campaigns focus on the consol¬ idation and spiritual repatriation of ethnic Russian minorities, based largely upon the rationalization of a shared identity. The striking similarities between the Ukrainian and Georgian experiences of identity driv¬ en disinformation campaigns typify post-Soviet state vulnerability to Russia’s revisionist resurgence. Both countries are linked to Russia by shared borders and a long history of political, economic and religious op¬ pression, not to mention occupation. Following the collapse of the former Soviet Union, Georgia, like gia, which turns on the question of national identity and language, illustrates the uniqueness of the Geor¬ gian case. The Russian narrative attacks on Ukraine suggest that to be Russian is vastly better than to be Ukrainian, to which the Ukrainians respond with de¬ fiant, inspirational messaging in support of their na¬ tional identity.7 When it comes to Georgia, however, Russia blurs and softens the boundaries of nation¬ al character — making it difficult to discern what it means to be Georgian. At the same time the Russian narrative does not, as a rule, denigrate Georgia’s na¬ tional identity, unlike its evident display of contempt for Ukraine’s political, social and economic attributes and consistent descriptions of its leadership as “na¬ tionalists, neo-Nazis, Russophobes and anti-Sem¬ ites.”8 The difference in narrative description of national identity is reinforced by language. In Ukraine, home to a sizeable ethnic Russian minority, propaganda ef- “Russia’s subtle appropriation of Georgian and Ukrainian national identities originates in Sovi¬ et efforts to control its minority populations. ” Ukraine, has attempted to repudiate its Soviet legacy and establish itself as a nation built on constitutional principles. Russia’s subtle appropriation of Georgian and Ukrainian national identities originates in Soviet ef¬ forts to control its minority populations. The USSR devalued the concept of ethnic identity and re¬ pressed the spread of ethnically motivated political nationalism byreplacing “national attachments” with generic (and artificial) values of solidarity and frater¬ nity.6 Removing ethnic singularity from the political lexicon enabled the Soviets to preempt radicalized discourse. Today Russia disinformation efforts in both countries are framed in a set of fuzzy assertions about a shared historical and religious heritage. These efforts include glossing over Stalin’s eviscera¬ tion of the Georgian Orthodox Church or the fraught history of multiple Russian imperial annexations of Ukrainian territories. The resultant narratives are laden with false claims to shared cultural and spiri¬ tual ethnicity. However, an important distinction between the Rus¬ sian disinformation campaigns in Ukraine and Geor- forts are almost entirely in Russian and consciously play up themes of ethnic isolation, cultural devalua¬ tion and feelings of disenfranchisement.9 The Rus¬ sian narrative actively portrays Russian minorities in Ukraine as victims of government sponsored vio¬ lence and encourages pro-Russian separatist efforts to discredit and even destabilize the ruling party. In Georgia, however, ethnic Russians only make up 1 .5 percent of the population, so the target for Russian propaganda efforts is almost exclusively Georgian.10 In fact, the pro-Russian voice in Georgia is Georgian.11 The Russian narrative does not attempt to appeal to a disenfranchised Russian minority, nor does it dis¬ credit current leadership. Rather it promotes pro Georgian sentiments — albeit on Russian terms — and lays the foundation for the claim that to be Georgian is to be Russian — or at least not European. RUSSIAN DISINFORMATION EFFECTS: VALUES The pro-Russian, anti-European narrative in the countries of the former Soviet Union asserts the ACPD I Crafting Resilient State Narratives in Post Truth Environments: Ukraine and Georgia 85 existence of a shared set of deeply held, historically based and culturally embedded shared values. Inev¬ itably, the Russian narratives portray the so-called “morally bankrupt West” in unflattering terms and play to deep-seated fears and prejudices held by tar¬ get audiences. The Russian narrative then offers the solace of a morally superior, pan-Slavic orthodoxy. The battle for the narrative in these countries plays out in a dynamic of opposing political, social and cul¬ tural norms. Much of the dominant imagery in the Russian nar¬ rative recalls past glories and recasts Russian impe¬ rial dominance in the region as examples of benign stewardship. These narratives also retell Georgian and Ukrainian history in terms that laud Russian mil¬ itary and political intervention. In reality, Russia has waged a series of wars in the region, using Georgia and Ukraine as a staging ground for its exploits and, in the process, absorbed territory and subjugated citizens. The Russian version of these events, how¬ ever, tells a story of benevolent protectionism rather than territorial occupation. In addition to retelling history, the Russian narra¬ tive in Georgia and Ukraine posits the existence of a closed community, hermetically sealed within the boundaries of greater Russia. For Georgia in partic¬ ular, the EU and the NATO represent a direct threat sovereignty and territorial integrity: “The EU Com¬ mission” will “define [Georgia’s] way of life, economic issues [and] policies.... This Commission is consid¬ ered the parallel government of Georgia.”12 Mean¬ while, as a consequence of a deepening relationship with NATO, Georgia will become “a transit territory... with a NATO camp training international terrorists.”13 Similarly the Russian narrative calls for the reunion of Russia and Ukraine through the embrace of their “shared” culture and history.14 The Russian disinformation narrative in Georgia, as in Ukraine, touts the primacy of the Orthodox Church, including the propagation of extremely con¬ servative attitudes about gender equality, sexuality and tolerance. This narrative is particularly powerful in Georgia, where prominent Georgian political and religious figures routinely claim that the West is in a “fight against Orthodox Christianity.” In other words, the only way that Georgia can be “saved” from a god¬ less West is by “partnering with Orthodox Russia.”15 In championing the morays of the Orthodox church, the Russian narrative frequently alludes to Europe’s “legalization” of “homosexuality, pedophilia and a perverse mode of life” by the West and claims that as part of the package of the EU Association Agreement, Ukrainian and Georgian citizens must embrace these corrupt values.16 Local, pro-Russian political leaders reinforce this homophobic narrative, conveying ag¬ gression toward and contempt for “LGBT people or their lifestyle or culture.” 17 RUSSIAN DISINFORMATION EFFECTS: MODELS OF GOVERNANCE The Russian narrative derives much of its illiberal, anti-democratic impetus from its national security strategy, which describes a series of politically moti¬ vated threats to Russian sovereignty: “The activities of... foreign and internation¬ al nongovernmental organizations, and fi¬ nancial and economic structures and also individuals, focused on destroying the unity and territorial integrity of the Russian Fed¬ eration, destabilizing the domestic political and social situation — including through in¬ citing ‘color revolutions’ — and destroying Russian religious and moral values.”18 The Rose and Orange Revolutions, which opened the door to increasing liberal democratic models of governance, posed an existential threat to Russia. No wonder that much of the Russian propaganda effort in Georgia and Ukraine attacks attempts to pursue further democratic reforms. A commentator on a pro-Russian Georgian language television station ar¬ gued, for example, that: “As long as the U.S. is in the region of the Caucasus, the dirtiness like the so-called Revolution of Roses, Orange Revolution and other troubles are very possible. They were invented to strengthen [the] Ameri¬ can regime.”19 By contrast, Russian propaganda narratives in Geor¬ gia and Ukraine assert Putin’s political invincibility, and the futility of resisting Russia’s might.20 Russian occupation of their territories reinforces this mes¬ sage. However, the anti NATO narrative plays out differently in the two countries. For Ukrainian audiences, Russia portrays NATO as a purely aggressive entity, bent on encircling and destroying Russia.21 By contrast, in Georgia, Russian propaganda draws on fears that 86 ACPD I Can Public Diplomacy Survive the Internet? the West has abandoned the Georgian people. The fact that Georgia has not been invited to join NATO, for example, becomes, in Russia’s nihilistic narrative, proof that the West does not have confidence in Geor¬ gia as a security partner. Instead, the Russians argue, it is better for Georgia to embrace its neutrality as a “non-Bloc” state.22 Finally, while Ukraine is already in active conflict with Russia, Georgia’s leadership fears the resurgence of open hostilities over South Osse¬ tia and Abkhazia and therefore has chosen not to run the risk of a public embrace of NATO. The subtext of the Russian narrative in Georgia is simple in its ma¬ levolence — Georgia needs Russia to restore its terri¬ torial integrity. The Russian anti-EU narrative in both countries plays on rural economic vulnerabilities, as well as residual nostalgia among older citizens who remain nostalgic for the relative security and stability of life in former Soviet Union.23 Unlike Ukraine, however, which re¬ tains a sizeable industrial base and technology sec¬ tor, Georgia’s economy is still largely agrarian and depends primarily on agricultural outputs. Accus¬ tomed to generous state subsidies and minimal prod¬ uct standardization under the old regime, Georgia’s farmers and food processors must now contend with complicated EU trade export regulations and the re¬ ality of open market competition under the EU Free Trade Zone. By contrast, Russia offers Georgia’s ag¬ ricultural producers immediate and unrestricted ac¬ cess to its own markets as well as membership in the less stringent Eurasian Economic Union.24 The Russians also regularly exploit challenges to the EU integration such as the Dutch failure to ratify the Ukrainian accession treaty or the long delay in Geor¬ gia’s entry into the EU visa liberalization program. The Russian version of events suggested that the West did not want to open its doors to Georgia’s “criminal ele¬ ments,” arguing that Berlin blocked a first-round de¬ cision to grant Georgian visa liberalization because it “feared a spike of crimes committed by Georgians in Germany.”25 Even after the implementation of the visa regime, it was all too easy to characterize as yet another instance in which Europe failed to behave ex¬ peditiously as Georgia’s advocate and partner. BUILDING A COUNTER-NARRATIVE Construction of a viable counter-narrative for Geor¬ gia and Ukraine in the post-truth environment begins with a clear, consistent and unified articulation of strategic priorities. Narrative resilience also requires coordination across government on messaging con¬ tent and dissemination, both internal and external. When appropriate, messaging should be synchro¬ nized with nongovernmental organizations (NGOs) and media sectors, at home and abroad. More indig¬ enous programming content should be developed to project a truly national identity and shared values. Existing legislation governing media and NGO licens¬ ing, ownership and financing must be improved and implemented for greater political transparency. Next, the Ukrainian and Georgian governments must deepen their understanding of target audienc¬ es needs and interests in order to develop effective message content. Visually compelling and easily un¬ derstandable representations of strategic interests and potential must appeal to external and internal audiences, as well as local and international opin¬ ion makers and journalists. Finally, the government, in cooperation with public and private sector insti¬ tutions, must build a regional and ultimately global network of journalists and news organizations that can support efforts to professionalize official media output and expand outreach efforts. Such networks would also facilitate the systematic investigation and exposure of the Russian state’s “weaponization” of in¬ formation. These steps can diminish Russia’s narrative domi¬ nance in Georgia and Ukraine. But the emergence of a truly viable counter-narrative requires that both countriesofferrealisticassessmentsofthecostsand risks associated with Euro-Atlantic integration. Un¬ met promises about NATO membership are easily ex¬ ploited, as are the stringencies of EU market access and production requirements. For Georgia, the development of a consistent count¬ er-narrative is further challenged by an apparent ambivalence within current leadership about the re¬ lationship with Russia and the West on all fronts — po¬ litical, economic and military.26 This, in turn “create[s] a feeling of ambiguity in society and contribute^] to Euro-skepticism.”27 Finally, the absence of official dis¬ course about linkages to liberal, pro-Western agenda creates a narrative void, allowing Russia to depict it¬ self as Georgia’s only viable ally and champion. At the same time, the Ukrainian narrative, in particu¬ lar, should focus on the representation of its potential as a regional economic and security partner, rather than a recap of its past victimization. The current fo¬ cus on the present crisis leaves no room for the pro- ACPD I Crafting Resilient State Narratives in Post Truth Environments: Ukraine and Georgia 87 jection of a better future. Moreover, if left unchecked, Ukraine’s healthy defiance of Russian aggression can easily morph into the projection of a form of ultra-na¬ tionalism not in keeping with its image as a tolerant, pluralistic nation. Before either country can arrive at a viable artic¬ ulation of strategic intent, they have some difficult questions to answer. First, they must decide on their respective identities in the post-cold war political sys¬ tem. Can Ukraine transition from a largely defensive account of its grievances to a positive projection of its regional potential? Will Georgia continue to be a beleaguered satellite of an imperious Russia or join the ranks of nations vying to exert power in a com¬ plex media environment? Then, both countries must live up to the values es¬ poused in their strategic narratives. Can Ukraine’s government make good on the civic momentum of the EuroMaidan? Will Georgia’s post-independence intent to become a liberal “beacon of democracy”28 be compromised by its domestic political, economic and security vulnerabilities? Finally, both countries must commit to a system of democratic governance consistent with stated values. But can they embrace “civic nationalism” and tolerance as long as the im¬ petus for illiberal “blood patriotism” — e.g. the annex¬ ation of their sovereign territories — remains?29 Ultimately, the sustainability of the Georgian and Ukrainian narratives in the global information space depends on the will and capacity to shift from a threat-driven reactive discourse to an opportuni- ty- based narrative that frames potential security and economic benefits in terms that resonate with target audiences. At all costs their narratives must avoid the tyranny of the stark rhetorical choice between “furi¬ ous Russia” and the “disgraceful West.”30 Effective persuasion lies in nuance, and the ability to commu¬ nicate the character and resilience of their national identities, values and models of governance. ENDNOTES 1. See, for example, Alister Miskimmon, Ben O’Loughlin, and Laura Roselle, Strategic Narratives: Communication Power and the New World Order, 2013. 2. Fukuyama describes “the emergence of a post-fact world, in which virtually all authoritative information sources [are] called into question and challenged by contrary facts of dubious quality and provenance.” Francis Fukuyama, “The Emergence of a Post-Fact World,” Project Syndicate, January 12, 2017. https://www.project-syndicate.org/onpoint/the-emergence-of-a-post-fact-world-by- francis-fukuyama-201 7-01 3. See, for example, Peter Pomerantsev and Michael Weiss, “The Menace of Unreality: How the Kremlin Weaponizes Information, Culture and Money,” a special report presented by The Interpreter, November 22, 2014 (New York: The Institute of Modern Russia), 12. http://www.interpretermag.com/wp-content/uploads/2014/11/The_Menace_of_Unreality_Final.pdf. See also Vivian S. Walker, “State Narratives in Complex Media Environments: The Case of Ukraine,” Case 331 , Institute for the Study of Democracy (Washington, DC: Georgetown University, 2015). 4. Georgi Kandelaki, “Georgia’s Rose Revolution: A Participant’s Perspective,” United States Institute for Peace Special Report, No. 167, July 2006, 11. https://www.usip.org/sites/default/files/sr167.pdf 5. See for example, Lasha Tugushi, Threats of Russian Hard and Soft Power in Georgia, GMF European Initiative - Liberal Academy Tbili¬ si, 2015 http://www.ei-lat.ge/images/doc/threats%20of%20russian%20soft%20and%20hard%20power.pdf and Tamar Kint- surashvili, Anti-Western Propaganda: Media Monitoring Report, 2014-2015, Media Defense Foundation, Tbilisi, Georgia, 2015. http://www. media-diversity.org/en/additional-files/documents/Anti-Western_Propaganda_Media_Monitoring_Report.pdf. See also Analysis of Russia’s Information Campaign Against Ukraine, NATO StratComm Centre of Excellence, 2014 http://www.stratcomcoe.org/analy- sis-russias-information-campaign-against-ukraine. 6. See Natalie Sabanadze, “Georgia’s ethnic diversity: A challenge to state-building,” in The Making of Modern Georgia, 1918-2012, Stephen F. Jones, ed.(Routledge, 201 3), 120-1 21, 123. 7. See for example, “Heaven Admits No Slaves,” trailer for documentary film, July 26, 2014. https://www.youtube.com/watch?v=bO- wi13a6Hd4. 8. Vladimir Putin, “Address by President of the Russian Federation,” The Kremlin, Moscow, March 18, 2014. http://en.kremlin.ru/ events/president/news/20603 9. Cited in Lasha T ugushi, Threats of Russian Hardand SoftPowerin Georgia, 1 2. Latvia — 26.2 percent, Estonia — 24.8 percent, Ukraine — 1 7.3 percent. 10. Ibid. 88 ACPD I Can Public Diplomacy Survive the Internet? 11. Author Interview with Giorgi Kldiashvili, Director of the IDFI, Tbilisi, Georgia, September 13, 2016. 12. From Russian sourced media cited in Tamar Kintsurashvili, Anti-Western Propaganda: Media Monitoring Report, 2014-2015, 27. 13. Ibid, 30. 14. See Analysis of Russia’s Information Campaign Against Ukraine, NATO StratComm Center. 15. Cited in Tamar Kintsurashvili, Anti-Western Propaganda: Media Monitoring Report, 2014-2015, 41 . 16. Ibid, 11. 11. Homophobia and Gender Identity, Media Monitoring Report 2014-201 5, Media Development Foundation, Tbilisi, Georgia, 10-11. http:// mdfgeorgia.ge/uploads/library/Homophobia-ENG-web%20(1).pdf 18. From Chapter IV, “Ensuring National Security,” in The National Security Strategy of the Russian Federation, #683, December 31, 2015. Cited in Tamar Kintsurashvili, Assessment of the EU Integration Communication and Information Strategy and its Action Plan of the Govern¬ ment of Georgia: Survey Report, Media Development Foundation, Tbilisi, Georgia, 2016, 12. http://ewmi-access.org/wp-content/up- loads/2016/06/Research-Publication-ENG.pdf 19. From Obiektivi TV cited in Tamar Kintsurashvili, Anti-Western Propaganda: Media Monitoring Report, 2014-2015, 22. 20. Author interview with Tamara Chergoleishvili, Publisher of Tabula, Tbilisi, Georgia, September 14, 2016. 21 . See Analysis of Russia’s Information Campaign Against Ukraine, NATO StratCom Center. 22. See “Kremlin’s Information War: Why Georgia Should Develop State Policy on Countering Propaganda,” Policy Paper, Institute for Development of Freedom of Information, Tbilisi, Georgia, August 22, 2016. https://idfi.ge/en/informational-war-of-krem- lin-against-georgia-the-necessity-of-having-state-policy-against-propaganda 23. Author interview with George Targamadze, Senior Fellow, Georgian Foundation for Strategic and International Studies (GFSIS), Tbilisi, Georgia, September 12, 2016. 24. Author interview with EU Delegation representatives, Tbilisi, Georgia, September 16, 2016. 25. “Hahn Confident on Ukraine, Georgia Visa Liberalization,” RFE/RL, September 3, 2016. http://www.rfer! .org/a/hahn-confident- ukraine-georgia-visa-free-2016/27965380.html 26. Author interview with Ambassador David Sikharulidze, Director, Atlantic Council of Georgia, Tbilisi, Georgia, September 13, 2016. 27. Tamar Kintsurashvili, Assessment of the EU Integration Communication and Information Strategy and its Action Plan of the Government of Geor¬ gia, Survey Report, 9. 28. See text of President George W. Bush Freedom Square speech in Tbilisi, Georgia May 1 0, 2005, BBC News, http://news.bbc. co.uk/2/hi/europe/4534267.stm 29. In articulating a “vision of Ukrainian identity for the 21st century,” Ukrainian politician Svyatoslav Vakarchuk argued: “We need to stop building a state based on blood patriotism and begin building a state based on constitutional patriotism. We shouldn’t be united by a common past, heritage, blood or appearance, but by a common set of values, lifestyles, rules and a constitution.” Or, as the Economist paraphrased it — there is a need to “replace ethnic nationalism with a more civic sort.” Economist, “Front man: Ukraine’s rock star politician,” October 22, 2016. http://www.economist.com/news/europe/21709067-pop-star-tries-help-coun- try-war-reinvent-itself-front-man 30. “This God-given land is attacked by tanks from furious Russia and by asses from the disgraceful West.” Cited in Tamar Kint¬ surashvili, Anti-Western Propaganda: Media Monitoring Report, 2014-2015, 25. ACPD I Crafting Resilient State Narratives in Post Truth Environments: Ukraine and Georgia 89 CC BY 2.0, USSR Political Poster from NichoDesign AMERICA’S STRATEGIC NARRATIVE AND A PATH FOR PUBLIC DIPLOMACY By Markos Kounalakis, Visiting Fellow, Hoover Institution, Stanford University The underlying strategic narrative driving American foreign policy priorities for the past 70-plus years, utilized time and again to encourage acceptance of Western values, is increasingly at risk. To be clear, this is not the result of one administration’s poli¬ cy preferences over another’s. Rather, consensus around certain values and goals — good democratic governance as a fundamental human right, the need to care for communities in crisis, a desire for equal rights for all, for example — is increasingly contested. This contestation isn’t because people have turned their backs on one another, are increasingly narcis¬ sistic, or mean spirited. Rather, it is because the sys¬ tems that this narrative support and normalize have failed to serve a large number of communities, both here in the United States and abroad. As crucial as new technology is to getting the right message out to the right audience, even the sleekest public diplo¬ macy campaign may be for naught unless we rebuild a domestic consensus regarding what we stand for as a nation and articulate how these values translate into foreign policy goals and priorities. A Greek friend, Yannis, always used to say that when he was younger, in the 1960s and 1970s, he was al¬ ways proud to see the Greek flag be the first to enter an Olympic stadium and, naturally, always rooted for his countrymen to win in competition. But he always had a second favorite nation: The United States. He felt it was patriotic and right to root for America to win over the dictators and demagogues who sat in the special seats reserved for the corrupt leaders of other nations. Greece first, America second was how he saw the world. And he was not alone. Times have changed. Yannis no longer has a second pick in these international competitions. He has not switched to rooting for Russia or Germany, but he no longer feels that emotional tug and viscerally driven, positive orientation towards America. Once again, he is not alone. In the quest for an “America first” policy, “America” may not last in the world’s popular imagi¬ nation. Traveling abroad, I’ve heard similar comments along the lines of, “The United States used to be better.” There was a time when regardless of American foi¬ bles — self-serving government interventions or, even, assassination plots — citizens in developed and devel¬ oping nations looked to the United States as a land of equality, political freedom and economic promise. They were willing, if not eager, to forgive American mistakes as temporary deviations from what was otherwise a noble and promising vision of the world. This instinct to forgive speaks precisely to the power of a strategic narrative and its centrality for effective public diplomacy. There is a prevalence of stories that have created a widespread domestic and international perception of an America that has finally come to terms with its own seemingly latent issues of corruption, conspira¬ cy and cynicism. Schadenfreude mixed with a confir¬ matory bias towards anti-Americanism, however, is a growing narrative as images, news stories and so¬ cial media amplify and focus on America’s perceived slights toward foreign leaders, policy failures and di¬ visiveness and hatred here at home. Put simply, the world can no longer buy the mystique, invest in the national brand and look to America as the beacon of freedom and defender of human rights if, as former Secretary of State Hillary Clinton once told me, “we no longer believe that story ourselves.” A RETURN TO BIG, BAD AMERICA Many have tried to attribute the three Cs of corrup¬ tion, conspiracy and cynicism to the American polit¬ ical and economic systems. Generally they have met with little success, in part because of our self-cor¬ recting and responsive political, economic and legal systems. Yet, the arrival of digital media platforms and a more sophisticated adversarial class utilizing bots and other digital tools may finally be muddying America’s messages with three Cs-themed criti¬ cisms, sometimes (but not necessarily) grounded in a modicum of fact. While a return to big, bad America is a contempo¬ rary aberration, it is not an historic exception. Recall the Reagan-era Kirkpatrick doctrine proposing a moderate approach towards friendly authoritarian states and a more confrontational posture contra totalitarian communist regimes. The same friend-foe calculus — sometimes described as a transactional 92 ACPD I Can Public Diplomacy Survive the Internet? Even the sleekest public diplomacy cam¬ paign may be for naught unless we rebuild a domestic consensus regarding what we standforasanationandarticulatehowthese values translate into foreign policy goals and priorities. ” approach — is being used today as the current admin¬ istration assesses its policies toward and the domes¬ tic stability of much of the Middle East. The recent praise of Egypt’s President Abdel Fattah el-Sisi was a calculated articulation of support for an authori¬ tarian regime that could be considered endangered by liberalization, with a concern that a rapid, uncon¬ trolled liberalization could again bring about a radi¬ cal, illiberal regime such as the Muslim Brotherhood government voted into power post-Mubarak. Similar levels of support have been expressed by the Amer¬ ican leadership for Turkey’s President Recep Tayyip Erdogan after the recent referendum granting him greater authority and power. This transactional approach, seemingly distant from a values-based approach that undergirded U.S. for¬ eign policy for so long, requires those who are deliv¬ ering public diplomacy goods and services to pivot is some significant ways. To start, public diplomacy needs to become more top-down, focusing on sys- temically aligning efforts with American national security priorities, while syncing with local, shared, policy goals. This approach may also mean less focus on national popular will and more on political elites, both demo¬ cratically elected and not, capable of leading public opinion toward consensual national alignment and support for American national security and econom¬ ic interests. In such an environment, civil society oppositional forces in foreign nations do not and, in the short run, may continue not to receive top-level access, support or recognition by U.S. government leaders. For ex¬ ample, in clear contrast to previous administrations, a recent visit to Moscow by Secretary of State Rex Tillerson did not include meetings with civil society or opposition leaders. Establishing this top-down approach in the early stages of a government may be necessary to reinforce both respect for national sovereignty and the building of trustworthy relations between senior governmental officials. It is understandable that an administration and of¬ ficials with minimal prior interactions at a govern- ment-to-government level would seek such closed- door and top-level interactions, eschewing pressures to test another nation’s tolerance or question its do¬ mestic approach or agenda. While understandable, this approach is also likely not sustainable. Domes¬ tic pressure from an American press and populace will demand interactions with opposition forces, civil society actors, minority groups and the cultural and educational institutions that make up the complex fabric of any society. This pressure, coupled with the long-term national interests of the United States, should expand the otherwise natural inclination to work more insularly and transactionally toward lim¬ ited goals and tight agendas. But even a short-term transactional approach will require engagementwithindividualsand institutions currently out of power or out of favor with governing regimes and ruling political elites. Public diplomacy must continue to build civil society actors and lead¬ ers, as they are likely to become the next generation of political leaders. Basic business logic and invest¬ ment strategies familiar to many within the adminis¬ tration require the hedging of bets, spreading of risk and investment in the future of any foreign govern- ACPD I America’s Strategic Narrative and a Path for Public Diplomacy 93 merit or administration. Shifts in political fortune and favor can be fickle and a strategy that makes medi¬ um- and long-term plays is a secure public diplomacy strategy. Public diplomacy emphases of the recent past have now changed and, either contemporarily or conse¬ quently, the message of liberal democracy has lost its shine, credibility and purchase. Given the espousal of a more transactional, efficient, business-centric, in- vestment-dominant and sovereignty-accepting polit¬ ical leadership in Washington, a more effective public diplomacy approach will need to be dramatically dif¬ ferent in order to be effective. Ifa Cold Warpublicdiplomacy model is helpfulincom- municating an approach to those formulating and implementing policy, then there is a cognate in the contemporary framework. The dominant Cold War adversary, however, is no longer a Moscow-based re¬ gime promoting a flavor of global communist ideolo¬ gy. The adversarial relationship with the Soviet Union has, in part, been supplanted by Beijing and Tehran (at the “peer-competitor” or “rising” state level) and “radical Islamic terrorism” (at the ideological level). As the Soviet successor state, Russia continues to challenge and threaten American interests — in par¬ ticular Russia’s still unfolding, yet the clearly formi¬ dable, information operations that are being investi¬ gated and assessed. But in the early days of the new administration, China and Iran were assigned great¬ er adversarial value. Given the relative power of these states and their ability to operate outside of a transactional frame¬ work, proxy conflicts may be primarily conducted be¬ tween the United States and China, with the potential for some alliance participation, and against Iran via a combination of military, economic and propaganda means aimed at containing and degrading the capac¬ ities of those states and their leadership structures. Recent April 2017 recalibrations regarding Rus- so-American relations may change some of the pub¬ lic diplomacy formulations, but to date the campaign and administration policy pronouncements and early administration actions dominate this analysis. Moving towards an updated Cold War II public diplo¬ macy framework and a rekindled reliance on a mod¬ ernized Kirkpatrick doctrine will mean that America respects a strong sovereignty approach towards friends and allies — regardless of their regime-type — and abstain from assertive policies and support for a civil society that can lead to dissatisfaction with gov¬ erning elites or regime change. In such a framework, however, adversarial regimes are subject to the full spectrum of American power and public diplomacy, from informational and educational programs to civil society infrastructural development for regime op¬ positional forces. Access to the populations of these adversarial regimes is the dominant constraint, of course, but these populations should be a primary American public diplomacy target. Regions, nations and regimes that are either neutral in their relations with the United States or are contested by regimes that are U.S. adversaries could also be a public diplo¬ macy priority. Access to the potential audiences of these contested places could face fewer constraints and their leaderships could be more open to straight transactional engagement, depending on whether the contested state is leaning towards the United States or towards an adversary. The following table represents this renewed public diplomacy approach in a resource constrained environment: New PD Matrix Sovereignty PD ECA Programs PD Civil Society Building Access Friends & </= Allies Contested + + (Leaning Friend) Contested + >/= +/- (Leaning Adversary) • + + +/- Adversary - + + - IRAN & ISIS ISIS is under significant military attack and Iran is under increasing political and military pressure, giv¬ ing support to domestic oppositional forces in both ISIS-held territory and within Iran. Public diplomacy activities should find greater receptivity in such an environment. In seeking a public diplomacy means for undermin¬ ing the ideological basis for supporting the current Iranian or ISIS structures, an emphasis on “Islamic Reformation” should factor in heavily. Two practi¬ cal target groups for this approach are women and youth, though they are by no means the only potential targets. 94 ACPD I Can Public Diplomacy Survive the Internet? Focusing on female empowerment as the primary public diplomacy goal within the Islamic-influenced world will allow the United States to maintain a mor¬ al component for American power and its liberation narrative. The demographic dominance of youth — and youth unemployment — in this region also offers a unique opportunity to focus public diplomacy efforts on in¬ culcating a dissatisfied demographic component with the tools and education to organize and con¬ front the inherent contradictions of these societies, such as religious elite privilege, regime elite power, social conformist demands, structural unemploy¬ ment realities, etc. Here, again, Cold War methods and practices are instructive in helping to introduce and reinforce messages and methods for organiza¬ tion and opposition to regional regimes. CHINA China’s current “Three Warfare” approach (using psychological, legal and media warfare) to policy priorities will be less effective in a U.S. -enforced, less permissive, global trade, maritime, military and political environment. In a more constrained, Ameri- an assertive public diplomacy approach with a more “one-size fits all” model. Given the previously outlined confrontational Ameri¬ can posture and constrained environment where an¬ ti-communist regime dissent will feel support, it will be in the U.S. interest to reinforce the more liberal factions of opposition and to seek support for a dif¬ fuse anti-Peoples Republic of China nationalist lead¬ ership base as well as expanded centers of regime opposition. Taiwan is a natural ally in this approach. Still vibrant forces — individuals and organizations in Hong Kong, in particular — should be targets of Cold War-style public diplomacy efforts. American diaspo¬ ra, visiting scholars and students, business interests with foreign investment in China, and other intersec¬ tional individuals and institutions should be cultivated and the targets of a public diplomacy program that can have impact. In terms of proxy practices favored during the Cold War, China has made significant inroads into both the African continent and in Latin America, using a successful combination of infrastructural invest¬ ment and regime support. Chinese public diplomacy efforts have promoted Chinese interests and used an “anti-colonial” narrative to undermine Western “ Public diplomacy has an opportunity to in¬ fluence a different international influencer class— the business, political, military and social elite. ” can-led, global environment, Chinese access to glob¬ al markets could become less open, free movement of shipping will encounter friction, and a state of “hu¬ miliation” could be perceived by any forced standing down of Chinese military (e.g. naval) and hybrid forc¬ es (e.g., off the North Korean coast). China poses a greater challenge in terms of popu¬ lation, geographic size and public diplomacy pene¬ tration. But there is a Cold War corollary here, too, including some natural advantages that did not exist during the Cold War. The most obvious one is that greater linguistic homogeneity allow for scalability of efforts in the recent past and into the current mo¬ ment. A constant and credible propounding of the failure of the “Washington Consensus” model of de¬ velopment — in particular in light of the 2008 global recession — has particular resonance. The promotion of the “Beijing Consensus” (sometimes sold in combi¬ nation with the “Singapore Model”) and the promise of the Asian Infrastructure Investment Bank (AllB), at a time when American aid and diplomatic resources are retrenching, could be a winning combination. Despite these inroads, there are some fairly straight¬ forward ways to mitigate China’s public diplomacy gains. Countering the Chinese narrative should be ACPD I America’s Strategic Narrative and a Path for Public Diplomacy 95 a public diplomacy goal. Dissatisfaction with the Chinese practice of demanding Chinese workforce deployment that accompanies Beijing-funded infra¬ structure projects (e.g., Ethiopia’s recently built rail¬ road system) and requires the on-going employment of Chinese nationals is already causing local conflict. Chinese finance terms for development is anoth¬ er source of irritation to host nations. Kickback schemes, regime-level corruption, media takeover and dominance, cultural imposition, societal elite fealty, and other characteristics of Chinese condi¬ tionality are ripe targets for a public diplomacy cam¬ paign to countering the current dominant Chinese relationships and accommodations. Cold War public diplomacy practices, again, are directly applicable, though demanding of localization in different parts of the African continent and in Latin America. Outside of the Islamic and Chinese front, public di¬ plomacy and diplomatic efforts in general, should be more concentrated in areas that are contiguous to those regions and that fall into what would be consid¬ ered a traditional geographic sphere of influence or ideological affinity zone. Other nations with natural or traditionally allied relations with the U.S. fall from any priority targeting and can be the recipients of whatever legacy public diplomacy programs exists to exchange cultural, educational and professional per¬ sonnel. Diplomatic relations should be maintained as close to status quo as is possible, all the while an¬ ticipating unplanned and occasional unpredictable social media attacks or policy pronouncements that could be counterproductive. Early indications are that there will be an active attempt on the part of ad¬ ministration cabinet members both to limit any po¬ tential damage, reinforce a policy continuity where it is aligned with previous administration foreign policy positions, and actively seek to privately reassure for¬ eign leaders and nations of the otherwise consistent application of policy and favor. In this environment, public diplomacy has an oppor¬ tunity to influence a different international influencer class — the business, political, military and social elite. Below are a few ideas: 1. One productive way to do public diploma¬ cy when policy priorities are unclear — or locally unpopular — is to go back to reliable projects and programs, such as empha¬ sizing sporting competitions and events. American prowess in athletics, whether the NBA or amateur athletics and Olympic sports, is the primary and popular way to achieve a level of cross-cultural penetra¬ tion. Sports are a favored means to pop¬ ular interest. An emphasis on American sports, training facilities, exceptional ath¬ letes and historical events can all build a positive image of a disciplined, admired and dominant America. The NBA is a global brand and marketing machine with a di¬ verse and global group of athletes playing in a uniquely American context. 2. Targeting authoritarian states that had fallen into disfavor during previous admin¬ istrations as a result of an emphasis on democratic transition and transparency. This is a moment during which countries where public diplomacy efforts have been unwelcome, rebuffed or countered in the past are now welcome. Hungary, Poland and the Philippines are only a few coun¬ tries that come to mind. Some authori¬ tarian countries that have been courted heavily by China in the last decade are now potentially in play. In the Middle East, a strongman posture that favors an elite and emphasizes America’s own, new — if partly romanticized — strongman leadership and projected, tough-guy posture has an op¬ portunity for penetration. When consider¬ ing public performance or appearance, a newer, more visible and heavier reliance on diplomatic security with military pres¬ ence to emphasize strength and perma¬ nence is favored to reinforce the big, bad nature of America. Think Death Star over Rebel forces. 3. Framing public diplomacy will be as import¬ ant as implementation of public diplomacy programming, especially as we move away from the “soft power” formulation. One potential frame: The ideal state for a more transactional public diplomacy environ¬ ment would be one that moves away from the “soft power versus hard power” con¬ struct and its inaccurate implications that “soft” is “weak” and instead move toward a public diplomacy deployment of “full spec¬ trum power,” where strong military, busi¬ ness, educational institutions and cultural products are forward leaning, leading and unrelentingly winning globally. Assuring 96 ACPD I Can Public Diplomacy Survive the Internet? and exhibiting that these institutions are the best globally would be consonant with the administration’s approach to policy and power projection. “Making America great again” includes — in fact, requires — “mak¬ ing American public diplomacy great again. Overall, the shift is for America to be feared and re¬ spected and away from an emphasis on openness and attraction. Achieving some of these program shifts in public diplomacy will be difficult given the traditions and culture of the majority of the American electorate and State Department training and practice to date. Some will be easier to achieve, such as the athletic and sporting component. Others will be more chal¬ lenging, such as adjusting public diplomacy messag¬ ing and practice to accommodate authoritarian lead¬ ers, their sycophants and supporters. In certain instances, public diplomacy profession¬ als may feel estranged from the new foreign elite that will be their audience and customer. A tighter relationship with the defense attache and with the commercial section will be required at post to make sure that public diplomacy is greatly aligned with U.S. commercial and business interests, as well as nation¬ al security priorities. This is a comfort zone for autocratic regimes and the transactional nature of the relationship is more de¬ fined, reliable and predictablefor leadership in these countries. Instead of playing towards a dissident elite, the target audience for the near future will be the ruling elite. The competition will be Chinese pub¬ lic diplomacy, in many cases, but may include a Russia both trained and successful at adversarial public di¬ plomacy practice. Finally, as this policy re-orientation may be tempo¬ rary, it is important not to burn all bridges. Maintain¬ ing relations with the non-ruling class will be helpful in the future. But taking advantage of the current mo¬ ment and in places where American public diploma¬ cy has either been shut out or unwelcome in the past, will allow for this administration’s policy priorities to find a more receptive audience. ACPD I America’s Strategic Narrative and a Path for Public Diplomacy 97 CONTRIBUTOR BIOGRAPHIES Amanda Bennett is a Pulitzer Prize-winning author, investiga- tivejournalist and ed itorand director of the Voiceof Ameri¬ ca. Previously, she was executive editor at Bloomberg News, where she created and ran a global team of investigative re¬ porters and editors. Bennett was editor of The Philadelphia Inquirerfrom June 2003 to November2006, and priorto that waseditorofthe Herald-Leaderin Lexington, Kentucky. Ben¬ nett served as a Wall Street Journal reporter for more than 20 years. A graduate of Harvard College, she held numerous posts at the Journal, including auto industry reporter in De¬ troit in the late 1 970s and early 1 980s, Pentagon and State De¬ partment reporter, Beijing correspondent, management ed¬ itor/reporter, national economics correspondent and, finally, chief of the Atlanta bureau until 1 998, when she moved to The Oregonian. She was also a contributing columnist for The Washington Post. Bennett shared the 1 997 Pulitzer Prize for national reporting with her Journal colleagues, and in 2001 led a team from The Orego/i/antoaPulitzerfor public service. MattChessen isa U.S. diplomat, technologistand American author who served in some of the most challenging assign¬ ments in the Foreign Service. He began his diplomatic career in Monrovia and subsequently served atthe U.S. embassy in Iraq, where he managed various disarmament programs. Returning to Washington, he led regional security and arms sales initiatives for the Bureau of Political-Military Affairs, covering Central Asia, Afghanistan and Pakistan. Chessen then served two years in Kabul advising NATO forces on the implementation of the Afghanistan Peace and Reintegration Program. He worked in Washington, D.C., in the OfficeofeDi- plomacyfrom 201 3-201 4, where he led the implementation of an innovative, open source, crowd-working platform called Open Opportunities. He subsequently served as the Coordina¬ tor for International Cyber Pol icy for the Bureau of East Asian and Pacific Affairs. ChesseniscurrentlyservingattheGeorge Washington University, where he researches the international implications of artificial intelligence astheState Department’s Science, Technology and Foreign Policy Fellow. Francis Fukuyama is Olivier Nomellini Senior Fellow at the Freeman Spogli Institute for International Studies (FSI), and the Mosbacher Director of FSI’s Center on Democracy, De¬ velopment, and the Rule of Law. He is professor (by courtesy) of political science. Fukuyama has written widely on issues in developmentand international politics. His book, The End of History and the Last Man, was published by Free Press in 1 992 and has appeared in over 20 foreign editions. His most recent book, Political Order and Political Decay: From the Industrial Revolution to the Globalization of Democracy, was published in September2014. Fukuyama received his B. A. from Cornell University in classics, and his Ph.D. from Harvard in political science. He was a member of the Political Science Depart¬ ment ofthe RAND Corporation, and ofthe Policy Planning Staff ofthe U.S. Department of State. Fukuyama is chairman ofthe editorial board of The American Interest, which he helped to found in 2005. Sam Ford researches how organizations are listening to, developing relationships with, telling stories to, and putting themselves in the shoes of the audiences they seek to reach. He currently consults on a range of projects, including act¬ ing as lead producer for a new Future of Work initiative in Kentucky for the MIT Open Documentary Lab. Previously, in 2015-2016, he ran the Center for Innovation and Engagement for Univision’s Fusion Media Group. Ford received his master’s degree from MIT. He isa research affiliateandconsultantwith MIT’s Program in Comparative Media Studies/Writing and teaches for Western Kentucky University’s Popular Culture Studies Program. In 201 3, he co-authored Spreadable Media (NYU Press), which has been translated into seven languages. Jeff Hancock is a professor in the Department of Communi¬ cation at Stanford University where he studies the psycholog- icalandinterpersonalprocessesinsocialmedia.Hisresearch specializes in using computational linguistics and experi¬ ments to understand how the words we use can reveal psy¬ chological and social dynamics, including studying the mental models people have regarding algorithms in social media. His research also explores how people use deception with tech¬ nology, from sending texts and emails todetecting fake online reviews. His TED Talkon deception has been seen over 1 million times and his research has been published in over 80 journal articles and conference proceedings with support from the U.S. National Science Foundation and the U.S. Department of Defense. Hisworkon lying andtechnologyhasbeenfrequent- lyfeaturedinthepopularpress, includingtheA/ew York Times, CNN, NPR, CBS and the BBC. Jonathan Henick, a member of the Senior Foreign Service, currently serves as the principal deputy coordinatorforthe Bureau of International Information Programs. He served previously as the acting deputy assistant secretary and di- rectorfor press and public diplomacy in the Bureau for South and Central Asian Affairs, where he was responsible for the conductofU.S.publicdiplomacyin13countries, including Af¬ ghanistan, Pakistan and India. He has also served overseas as the counselor for public affairs in Turkey, the deputy chief of mission in Timor-Leste, as well as in other positions in Azer¬ baijan, Turkey, Portugal and Uzbekistan. He has worked as a publicdiplomacyfellowand professor at George Washington University’s School of Media and Public Affairs and as a visiting research fellow and diplomat-in-residence at the East-West ACPD I Contributor Biographies 99 Center in Honolulu, Hawaii. He has received the Public Diplo¬ macy Alumni Association Achievement Award, as well as in¬ dividual Superior Honor Awards from the State Department. Henick speaks Russian, Portuguese, Turkish and Azerbaijani, and holds a Ph.D. in political science from the University of Hawaii. Tim Hwang is a partner at Robot, Robot & Hwang, a lawfirm and technology consultancy focusing on experiments at the intersection of legal and computer code. He leads its initiative seeking to develop general principles and common frame¬ works to guide policymaking as intelligent systems emerge and become increasingly ubiquitous in a variety of arenas including capital markets, warfare, medicine, transportation and social lifeatlarge. Markos Kounalakis is a visiting fellow at the Hoover Institu¬ tion at Stanford University and a seniorfellowatthe Center for Media, Data and Society at Central European University in Budapest, Hungary. He is a presidential^ appointed mem- berof the J. William Fulbright Foreign Scholarship Board, president and publisher emeritus of the Washington Month¬ ly, and writes a foreign affairs column for The Sacramento Bee and McClatchy-Tribune News. Kounalakis received his Ph.D. from Central EuropeanUniversityin2016andiscurrent- ly writing a book on the geopolitics of global news networks. He has written three books: Defying Gravity: The Making of Newton (Beyond Words Publishing, 1993); Beyond Spin: The Power of Strategic Corporate Journalism (coauthor, Jossey- Bass Publishers, 1999); and Hope is a Tattered Flag: Voices of Reason and Change for the Post-Bush Era (PoliPointPress, 2008). Kounalakis serves on the Board of Councilors at the University of Southern California’s (USC) Annenberg School for Communication and Journalism; and the Board of Advi¬ sors at USC’s Center on Public Diplomacy (CPD). Ethan Porter is an assistant professor at George Washing¬ ton University in the School of Media and PublicAffairs. He received his Ph.D. in political science from the University of Chicago in 2016. His research interests include public opinion, political communication, political psychology and experimen¬ tal design. Porter has received grants from the National Sci¬ ence Foundation and the Omidyar Network. Shawn Powers serves as the executive director of the Unit¬ ed States Advisory Commission on Public Diplomacy. He has a Ph.D. from the Annenberg School for Communication and Journalism atthe University of Southern California (USC) and more than a decade of experience working at the nexus of publicdiplomacy,developmentand national security. Powers researchesthegeopoliticsof information andtechnology and published (with Michael Jablonski) the award winning The Real Cyber War: A Political Economy of Internet Freedom (The Universityof Illinois Press, 2015). Hehasover40 publications in academic and mainstream outlets, including The Washing¬ ton Post, Guardian and Huffington Post. His research has been supported by grants from the British Council, U.S. Depart¬ ment of Defense, U.S. Department of State, European Com¬ mission, Knight Foundation, Open Society Foundation, and U.S.InstituteforPeace.Hehasalsoreceivedfellowshipsfrom the London School of Economics, University of Pennsylvania, and Oxford University. Laura Roselle is professor of political science and internation¬ al studies at Elon University whereshe is currently a senior faculty fellow. Roselle holds degrees from Emory University (math/computer science and Russian) and Stanford Univer¬ sity (Ph.D., political science). She has served as president of the International Communication Section ofthe International Studies Association and ofthe Internet Technology and Pol¬ itics Section of the American Political Science Association. She is the author of Media and the Politics of Failure: Great Powers, Communication Strategies, and Military Defeats (Pal- grave, 2006 &2011), and, with co-authors AlisterMiskimmon and Ben O’Loughlin, of Strategic Narratives: Communication PowerandtheNewWoridOrder(Roui\edge, 201 3) and Forg¬ ing the World: Strategic Narratives & International Relations (University of Michigan Press, 2017). Roselle is co-editor of the journal Media, War and Conflict, and co-editor ofthe book series, Routledge Studies in Global Information, Politics and Society. She won the 2017 Distinguished Scholar Award from the International Communication Section ofthe International Studies Association. Jason Stanley is the Jacob Urowsky Professor of Philosophy atYaleUniversity. BeforecomingtoYalein201 3, hewasdistin- guishedprofessorinthe Department of Philosophy at Rutgers University. He has also been a professor at the University of Michigan (2000-2004) and Cornell University (1995-2000). His Ph.D. was earned in 1 995 at the Department of Linguis¬ tics and Philosophy at MIT (Robert Stalnaker, chair), and he received his bachelor’s degree from the State University of New York at Stony Brook in 1 990. Stanley has published four books— two in epistemology, one in philosophy of language and semantics, and one in social and political philosophy. His latest book, How Propaganda Works, was published by Princ¬ eton University PressinMay2015. Itwasthewinnerofthe 2016 PROSE award for the subject area of philosophy. In 2015, Stanley received a doctor of humane letters, honoris causa, from Binghamton University. Vivian S. Walker is a professor of national security strategy at the National War College in Washington, D.C. Previously she served as a professor of strategic and security studies at the National Defense College ofthe UAE and a visiting professor atthe Near East South Asia Centerfor Strategic Studies at the National Defense University. In her career with the State Department, she twice served as a deputy chief of mission 100 ACPD I Can Public Diplomacy Survive the Internet? (Croatiaand Armenia), twice asan office director (Southeast¬ ern Europeand EUR’sOfficeof Pressand Public Diplomacy), a public affairs officer (Kazakhstan), a cultural affairs officer (T unisia), and an information officer (Haiti). Other assign¬ ments include a professorship at the National War College, the regional border coordinator in Afghanistan, and a fel¬ lowship on the U.S. Atlantic Council, where she led the first interagency discussion on public diplomacy post 9/11. Walker graduated from Georgetown University’s School of Foreign Service and received a doctorate from the University of Chi¬ cago. She speaks French, Russian and Croatian. Ryan E. Walsh is senior advisor for d igital product at the De¬ partment of State’s Bureau of International Information Pro¬ grams (IIP) where he is responsible for the planning, produc¬ tion and optimization of digital contentfor use at post. Ryan joined 1 1 P with over 1 0 years of private sector experience in digital-first content development, newsroom editorial strate¬ gy a nd social network data analysis. Most recently, he hel ped launch the data-driven online news startup Vocativ. In its f i rst yea r as a site , Vocativ ach ieved a n average of 5 m i 1 1 i on unique monthly views and the Scripps Howard Foundation recognized Ryan as a finalist for the national award in digital innovation. He previously worked in crisis communications forGoldmanSachsandisadigital advertising agency veteran of McCann WorkGroup. He holds a master’s degree in global affairs from NYU and a bachelor’s degree in historyfrom Prov¬ idence College. Bruce Wharton is the acting under secretaryfor publicdiplo- macy and public affairs at the U.S. Department of State. Am¬ bassador Wharton served as the principal deputy assistant secretary in the Bureau of African Affairs from 2015-2016. Prior to that he served as the U.S. ambassador to Zimbabwe from September 201 2-November 201 5. He has also served as the Bureau of African Affairs deputy assistant secretary forpublicdiplomacy,Africanaffairsdirectorofthe Office of Public Diplomacy and Public Affairs, and deputy coordinator of the Department of State’s Bureau of International Informa¬ tion Programs. From 2003-2006 he was the deputy chief of mission atthe U.S. embassy in Guatemala. Wharton received Superior and Meritorious Honor Awards from the Depart¬ ment of State and the U.S. Information Agency, and was the 2011 recipient ofthe Edward R.MurrowAwardfor Excellence inPublicDiplomacy.HeisagraduateoftheUniversityofTexas in Austin and speaks Spanish and German Sam Woolley is the director of research for the Oxford In¬ ternet Institute’s European Research Council (ERC)-funded Computational Propaganda Project. Sam specializes in the study of automation and politics, with special interests in po¬ litical communication and science and technology studies. His workon botsand publicopinion has been published in several academic journals and collections. For his research, he has been featured in publications such as Wired, Fast Company, The Washington Post, The Economist and Bloomberg. Sam is aPh.D.candidate(ABD)attheUniversityofWashingtoninthe Department of Communication, and afellow at Jigsaw, the In¬ stitute forthe Future, and theTechPolicy Lab atthe University of Washington. ACPD I Contributor Biographies 101 www.state.gov/pdcommission