Jump to ContentJump to Main Navigation
Network PropagandaManipulation, Disinformation, and Radicalization in American Politics$

Yochai Benkler, Robert Faris, and Hal Roberts

Print publication date: 2018

Print ISBN-13: 9780190923624

Published to Oxford Scholarship Online: October 2018

DOI: 10.1093/oso/9780190923624.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 23 October 2019

Are the Russians Coming?

Are the Russians Coming?

Chapter:
(p.235) 8 Are the Russians Coming?
Source:
Network Propaganda
Author(s):

Yochai Benkler

Robert Faris

Hal Roberts

Publisher:
Oxford University Press
DOI:10.1093/oso/9780190923624.003.0008

Abstract and Keywords

This chapter examines the evidence supporting the claim that Russia mounted sustained and significant information operations in the United States. It finds that the evidence of Russian inference is strong but that the evidence of its impact is scant. The documented efforts of Russian interference typically entail piling onto existing debates and seeking to exacerbate existing social divisions. This chapter emphasizes that it is critical not only to understand that Russian propaganda efforts occurred but also to evaluate the effectiveness of these operations. If the biggest win for Russian information operations was to disorient American political communications then overstating the impact of those efforts actually helps consolidate their success. But it is important not to confuse the high degree to which Russian operations are observable with the extent to which they actually made a difference to politically active beliefs, attitudes, and behaviors on America.

Keywords:   presidential election, Russia, propaganda, hacking, social media, sockpuppets, bots, cyborgs, useful idiots, right-wing media

BY EARLY 2018 only the willfully blind or complicit could continue to reject the proposition that Russia has been running information operations for years aimed at intensifying political divisions and weakening democracy in the United States and Europe. From diverse perspectives and efforts, accounts by journalists and academics, intelligence agencies in the United States and Europe, and the special counsel investigating relations between the Trump campaign and Russia have supported this finding. Even the Trump-supportive House Intelligence Committee investigation confirmed that Russia had mounted cyberattacks on American targets and used social media to sow discord during the 2016 election. A weaker European Union, and a more fractured U.S. polity and inwardly-focused U.S. foreign policy, would leave Russia with greater leeway to pursue its own foreign policy goals, and a steady stream of news about Western democracies in disarray offers legitimacy for Vladimir Putin’s own brand of strongman illiberal populism. If we were persuaded that Russian propaganda and interference were the primary reason that the United States elected a president so congenial to Russia’s interests; that the third largest economy in Europe, the United Kingdom, had voted to depart the Union; that Spain and Catalonia came to the verge of splitting; or that Italy’s Euroskeptics won a majority of the seats in Italy’s parliament, we would be observing the most dramatically successful attack on democracy since the end of World War II.

Three primary sources of skepticism, or at least caution, about the Russian interference hypothesis emerged during the year after the election. First, supporters of President Trump resisted the implication that Trump’s election was a victory for Russian information warfare not the American people. Second, veteran watchers of the controversy over weapons of mass destruction and intelligence on Iraq, and the media groupthink failures that (p.236) accompanied them, took the interagency consensus on Russia with a grain of salt. Both of these lines of skepticism were on display in Chapter 5: the Fox News interventions for the former, and the Glenn Greenwald articles for the latter. The third form is standard academic working skepticism. This is simply the result of applying rigorous standards of proof to claims and assessing the stronger and weaker aspects of claims by these standards. In this chapter we apply this standard and explain why we are persuaded by the weight of the evidence that there was a sustained Russian effort; but we offer reasons for caution around some of the more expansive narratives—that the social media environment was overrun by Russian bots or that street protests in America were fomented to a significant degree by Russian agitation. We emphasize the difference between proof of the existence of Russian efforts and proof of their impact and suggest that the evidence of impact is less clear. In particular we are guided by the observation that the primary goal of Russian disinformation is to instill a sense that “Nothing is True and Everything is Possible,”1 or to “dismiss, distort, distract, and dismay.”2 If that is the case, then to overstate the prevalence and effect of Russian attacks is to aid their success. Just as terrorism succeeds most when it evokes an overreaction and causes a society to respond from fear and anger rather than calculation, so too will Russian active measures have their largest effect through evoking a harmful autoimmune response from the countries under attack.

Our primary contribution here is to examine the interaction between the asymmetric architecture of the American media ecosystem that we discovered in our own work and the presence of the Russian efforts. In a nutshell what we observe is that Russian efforts take advantage of existing fissures and pathways. The differences between the right-wing media ecosystem and the ecosystem occupied by the rest, from the center to the left, make the former more susceptible to propagandist efforts than the latter. The insular, domestically produced network of sites and social media diffusion networks that traffic in politically motivated falsehoods, coupled with the persistent attacks on mainstream media and other evidence-based institutions and expertise, have made the right wing of the American media ecosystem more susceptible to penetration, less resilient, and less capable of self-correction. When Russian propaganda efforts are consistent with right-wing American framings and beliefs, these falsehoods are able to insert themselves, propagate, and gain credence in the right-wing media ecosystem. By contrast, similar efforts aimed to leverage left-wing biases have to overcome the basic checks provided by a media ecosystem inhabited by professional-norms-oriented media outlets.

(p.237) Russian Origins

Vladimir Putin’s first term in office from 2000 to 2008 was marked by a highly restrictive policy toward mass media but permissive policy toward online communication. Internet penetration in Russia was low, and the reputational costs of a significant crackdown on a symbolically important but politically marginal medium likely outweighed the perceived political risks associated with letting the internet flourish. By the end of Putin’s first two terms, LiveJournal had become the dominant blogging platform, and a report authored by Bruce Etling, one of us (Faris), and several other of our colleagues concluded “the political blogosphere appears to remain a free and open space for Russians of all political stripes to discuss politics, criticize or support government, fight corrupt practices and officials, and to mobilize others around political and social causes.”3 It was into this relatively open, highly contested networked public sphere that Dmitry Medvedev entered, earning the nickname “Blogger-in-Chief” from one observer.4 In support of his activity, Medvedev’s government began to redirect youth movements originally developed to counter any Russian “color revolution” efforts on the streets to online activity, and trolls and bots made their first observed appearance in Russian online debates.5

The major change, however, coincided with Putin’s return to office in 2012 and in response to the most sustained and forceful protests since the implosion of the Soviet Union. A combination of changes in ownership and direct legal pressure forced online platforms Yandex, Vkontakte, and LiveJournal to suppress oppositional voices, and media critical of the government were forced to change leadership, often leading reporters to emigrate and set up shop abroad. Moreover, the wide-open, commercial nature of the Russian internet before 2012 gave birth to companies highly proficient in search engine optimization and commercially oriented trolling. These capabilities became the foundation of the new state-managed online information campaigns against opposition leaders.

In 2013 independent Russian investigative journalists published the first reports on the Internet Research Agency (IRA), which would ultimately be indicted by Robert Mueller in 2018. Registered in 2013, the IRA would become a troll factory, employing paid agents to shape online conversations. Russian investigative reporters critical of the government went undercover and reported repeatedly on the agency from 2013 to 2015.6 Looking at the research on online communications in Russian politics from the first decade of the twenty-first century, it seems quite clear that the capabilities deployed (p.238) against the United States in the 2016 election cycle and beyond were still in their infancy by the time they were deployed. There was one investigative journalism piece by Russian expats as early as 2003 observing patterns of information manipulation by “Internet brigades” within the Russian media ecosystem. That report was also the basis of the first academic English-language description of Russian “third generation controls” on internet discourse in 2010 by Deibert and Rohozinski.7 Deibert and Rohozinski identified coordinated campaigns by “ ‘Internet brigades’ to engage, confuse, or discredit individuals or sources. Such action can include the posting of prepackaged propaganda, kompromat, and disinformation through mass blogging and participation in Internet polls, or harassment of individual users, including the posting of personal information.”8 The most publicly prominent voice raising the alarm about this class of attacks on democracy was Evgeny Morozov in The Net Delusion in 2011, who offered it as a frontal assault on the then-prevailing wisdom that the internet democratizes political discourse.9

The early investigative journalism and academic work on the Russian internet was complemented by national security research following the Ukranian crisis of 2014.10 Reports from researchers studying Russian national security, including think tanks in Central and Eastern Europe and some in NATO, began to evaluate operations in Russia and its near sphere, from Georgia to Ukraine, Moldova, and beyond. By March 2015 the European Council was concerned enough to form the European Union’s East StratCom Task Force to deal with the dangers of Russian information operations in the “Eastern Partnership” countries—Armenia, Azerbaijan, Belarus, Georgia, Moldova, and Ukraine. This body of research precedes the U.S. election and identified strategies remarkably congruent with those observed in other countries by diverse researchers who are not all on the same page. For skeptical Americans concerned that the Russia story is a groupthink overreaction to the election, this body of research should offer some increased confidence that it is real. Perhaps, as some suggest, present Russian doctrine merely updates Soviet “active measures.” Perhaps it reflects adaptation and extension of strategies that Russian military and intelligence leaders believe the United States used to undermine Russia in its own sphere of influence. The most obvious trajectory is that what we saw in Russian operations in America is merely an extension of the strategies that the Russian state developed to quell domestic political debate after the 2011–2012 protests. Whatever its origin, the sum of the extensive Russia-centered literature is to document that present Russian doctrine and practice, at home, in its near periphery and increasingly (p.239) in the European Union, involves extensive information operations of roughly the form observed in the United States during the election period and since. These techniques include breaking into computer systems to obtain compromising information and releasing it in public; creating false accounts on social media to propagate opinions and reports; leveraging the affordances of social media advertising platforms; and combining white propaganda on RT and Sputnik with gray and black propaganda sites and social media accounts to disseminate and add credibility to false information and propagate destabilizing disinformation.

Hacking and Doxxing

Beginning a week after Hillary Clinton formally secured a majority of the Democratic delegates to the convention, and culminating in a big WikiLeaks email dump on the Friday before the Democratic National Convention opened, someone began to release batches of emails on a site called DCLeaks, emails obtained by hackers from the email system of the Democratic National Committee (DNC). The crux of the most damaging emails, particularly those released by WikiLeaks on the eve of the Democratic National Convention on July 22, 2016, showed that DNC leadership had strongly favored Hillary Clinton over Bernie Sanders, despite the organization’s commitment to neutrality among the primary candidates. This forced the resignation of then-DNC Chair Debbie Wasserman Schultz.11

Hacking into opponents’ email and computer systems, obtaining compromising information, and then “leaking” it back out to the public as part of an online propaganda campaign is a technique Russian propaganda efforts have used repeatedly in Russia and abroad.12 Immediately after the first DNC breach, CrowdStrike, the security firm hired by the DNC to investigate the breach, came out with a report claiming that the hack had been perpetrated by two Russian groups, Fancy Bear and Cozy Bear. One was allegedly associated with the GRU, Russian military intelligence, the other was possibly associated with the FSB (successor to the KGB).13 Over the next few months, other information security firms and expert groups such as FireEye and ThreatConnect added their own independent assessments, concluding that the digital forensic evidence supported the proposition that the DNC hack was perpetrated by Russian operations.14 In October and December of 2016, the Department of Homeland Security and the FBI released joint statements asserting that the DNC hack was likely directed by the Russian government and that the attacks and (p.240) subsequent leaks were “consistent with the methods and motivations of Russian-directed efforts.”15

On January 6, the Office of the Director of National Intelligence (ODNI) released a joint assessment by the FBI, CIA, and NSA that included a broader and more detailed set of assertions.16 This assessment attributed the DNC and Podesta email hacks to Russian state actors (although on this question the report was heavy on “trust us” and light on forensic details). It also claimed that these were part of a broader propaganda and social media campaign (again, though, focused on RT rather than any more exotic efforts). And it explicitly rejected the idea that Russia had tried or succeeded in any way in tampering with the 2016 vote-tallying process. Pushing back on the findings of this report was a central feature of the Seth Rich story we noted in Chapter 5. We return to that theme here in the context of describing white and gray propaganda. For now, we accept the consensus of diverse commercial information security firms, technology trade press investigations, mainstream media reports, and the consensus assessment of the U.S. intelligence community.

Social Media Influence: Sockpuppets, Bots, Cyborgs, and Ads

The indictment issued by Robert Mueller against the Internet Research Agency, Yevgeny Prigozhin, and other Russian nationals offered detailed allegations regarding the pattern of Russian intervention, combining false accounts on Facebook and Twitter, bots, strategic targeted advertising bought under false identity, and mobilizing real-space protests by unwitting Americans.17 The core strategy, according to the indictment, was to increase disaffection, distrust, and polarization in American politics. The practical targeting seemed intended to boost enough of the white-identity vote to drive voters to the polls in states that made the difference, to suppress enough of the black vote, and to suppress enough of the left vote or divert it to Green Party candidate Jill Stein, to tip the election. Given the baseline purpose of undermining trust in democratic institutions, however, simply sowing these divisive themes would have been a perfectly reasonable consolation prize. There were two core claims in the indictment. First, it claimed that the defendants had infiltrated social media networks through the use of false accounts, whether fully automated (bots), humans masquerading as someone other than who they are (sockpuppets), or some combination of the two (cyborgs). Second, it claimed that they bought targeted political advertising online under false accounts and presented themselves as authentic American political activists to spread divisive narratives, organize real-world protests, (p.241) and push stories that undermined Hillary Clinton and supported Donald Trump. We take this indictment not as proof but as one piece of evidence among many pointing toward the reality of Russian interference in the election. We will use its precision and confidence later in this chapter to raise questions about the likely actual impact on the election and the American public sphere more generally.

The indictment’s allegations are congruent with independent reports by investigative journalists, disclosures by Facebook, Twitter, and Google in their testimonies to the House and Senate intelligence committees, and academic studies of Russian operations. Following the first round of committee hearings, Facebook issued a report in April 2017 on “information operations” that was widely read to cover Russia’s activities, although it avoided saying so. In anticipation of the second round of congressional hearings, Facebook released an updated report,18 as did Twitter.19 Facebook, Twitter, and Google returned to testify in late October, when the most extensive disclosures occurred. Democrats on the House Intelligence Committee released a sample of ads that had been shown on Facebook and that, according to the company, had been purchased by agents of Russia. The ads disclosed various groups masquerading as American activists on both the right, such as Defend the 2d, Secured Borders, and Stop A.I. (where A.I. stood for “All Invaders,” not “artificial intelligence”) and the left, such as Blacktivists and LGBT United. These releases included methods clearly intended to instigate street clashes, such as publishing ads for opposing protests on the same day across the street from each other, one called “Stop Islamization of Texas” and the other called “United Muslims of America.” Facebook’s testimony reported that the Internet Research Agency campaign had published 80,000 posts that were shown directly to over 29 million people.

A team led by Young Mie Kim of the University of Wisconsin leveraged the Facebook groups identified by the minority report of the House Intelligence Committee to produce the best study of Facebook advertising by, among others, suspicious Russian groups. They found that 5.6 percent of the Facebook ads served to their large, nationally representative sample of subjects were from one of these sites, and their work obviously identified only a lower bound of Russian intervention using purchased ads.20 Facebook also testified that Russians had used Instagram to share over 120,000 photos from 170 accounts. And committee Democrats released a list of over 2,700 Twitter handles that have since become the canonical set in efforts to study Russian influence on Twitter. Twitter claimed that the handles had posted over 130,000 tweets during the election period. Beyond these, Twitter (p.242) testified that the company had identified 36,000 Russia-linked bots that had shared 1.4 million tweets that received about 288 million views. Google, for its part, testified that it believed the Internet Research Agency had bought ads on YouTube, ran 18 channels, and uploaded over 1,100 videos, which they reported as having received very few views: a total of 309,000 views spread across the 1,100 videos. All three companies emphasized that, while the absolute numbers seem large, they represent tiny slivers of the total amount of posts, tweets, videos, and other engagements and uses of their respective sites.21

Both before and after these reports, journalists reported a series of stories fleshing out these techniques. A report in the Daily Beast, for example, was the first to document a real-world anti-immigrant protest in Twin Falls, Idaho. The rally was organized using Facebook events by the Russian-controlled site Secured Borders.22 CNN similarly reported and offered details of the incendiary efforts of Blacktivist, a month before the details about the account emerged in congressional testimony. The Wall Street Journal reported that four of these highly divisive accounts, Secured Borders, Blacktivist, Heart of Texas, and Being Patriotic, had between them nearly a million followers.23 The New York Times tried to track down “Melvin Reddick of Harrisburg Pennsylvania,” who used Facebook to promote DCLeaks, the site initially used to tease the DNC emails before the broader release on WikiLeaks. They were unable to locate anyone in the United States who matched the profile, but a Brazilian site took interest and with help from their readers located the man, a Brazilian national, whose picture was used in the account.24

As a report from the Atlantic Council’s Digital Forensics Labs makes clear, however, the most extensive and in-depth reporting came from independent Russian journalists, some working abroad for fear of retaliation, who uncovered much of this activity.25 Polina Rusyayeva and Andrei Zakharov, journalists at RBC, a media group that has published stories critical of business interests close to Valdimir Putin and come under pressure from the Russian government for it,26 published the most comprehensive leaked set of Facebook, Twitter, and Instagram accounts run by the Internet Research Agency in the U.S. election. These included accounts aimed at African American audiences, such as Blacktivist or DON’Tshoot; accounts aimed at anti-immigrant sentiment, such as Secured Borders and Stop all Invaders; at Republicans, such as TEN_GOP and Tpartynews; and at police officers, such as Back the Badge or Pray4Police; and so forth.27

Some more detail on the commonly-used terminology is useful. “Bots” refer to fully automated accounts or scripts that simply share, tweet, or retweet (p.243) news headlines, with or without links, retweet other accounts, or exhibit similarly simple actions in high volumes with minimal human supervision. “Sockpuppets” usually refers to fully human accounts that pretend to be people they are not. We prefer “sockpuppets” to “trolls,” because the “trolls” term has been used in two quite different meanings in the debate. In long-standing studies of 4chan and other internet countercultures, the term had quite distinct meanings that were more nihilist than programmatic. As three of the leading scholars of these hacking and trolling subcultures argue, by 2017 “trolling” in American usage had come to mean too many things in too many contexts to continue to be a useful analytic category.28 In Russia-centered studies, by contrast, “trolls” means human participants paid to intervene in an online conversation to shape it in favor of whoever is paying them. That was the Russian usage, identified in Russia, by the Novaya Gazeta report that first outed the Internet Research Agency.29 We prefer “sockpuppet” to “human” because we think it important to separate out instances where human beings are presenting their own real persona, even if paid, and those where an individual takes on one or more persona that they manage in a fully human process but present to the outside world as someone other than who they are. We mean this to cover individuals paid to post repeatedly, captured by the “50 Cent Army” metaphor applied to both China and Russia, whether they do so by controlling one account or several. Increasingly these actors use scripts and supervised automation to increase their effectiveness. Accounts that use these practices are increasingly described as “cyborg” by social media researchers.30

White Propaganda, Gray Propaganda, and Useful Idiots

There are several other major pieces of the Russian interference puzzle: the role of the state-owned RT and Sputnik in developing and reporting stories targeted at specific countries; the role of apparently local media outlets that are in fact run by Russian agents; and the role of genuinely local media that propagate Russian-developed narratives, wittingly or unwittingly, either to push a political agenda or to attract readers and advertising dollars. The latter are the “useful idiots.” In Cold War–era propaganda studies, “white” propaganda referred to communications by an outlet acknowledged to be of one party or another, where the information was more or less accurate but framed to favor that party. “Black” propaganda referred to spreading lies and deception through falsely described sources. “Gray” propaganda was somewhere in between.

(p.244) The January 6, 2017, ODNI report focused heavily on RT as a major source of propaganda. The report particularly emphasized a number of statements that RT editor-in-chief Margarita Simonyan made to various Russian and foreign outlets. Those statements explicitly described RT’s role as a state-funded outlet to serve as an important part of Russia’s defense system in the information wars.31 In our own studies RT was a moderately visible site during and after the U.S. election, shared more by Trump followers than Clinton followers, by a 3:2 ratio. The relatively low direct visibility of RT, however, understates the impact of the site through repetition of its stories on other sites in the network. Some of those propagandist sites are masked Russian outlets (gray propaganda), and some are pawns who repeat the Russian line for their own purposes.

Infowars, the eighth-most tweeted site on the right, repeatedly republishes RT content. A BuzzFeed investigative report found that Infowars had republished over 1,000 stories with an RT byline between 2014 and 2017, without RT’s permission.32 Analyzing the text of stories in our own 2017 set, we found 431 stories in which a sentence of at least 32 words on Infowars was identical to one found on RT. Those stories account for about 3 percent of the Infowars stories in our 2017 dataset. The stories generally either simply republish RT content or quote and link to the site. Sean Adl-Tabatabai, the author of YourNewsWire.com, which has risen to a position of significant visibility in the right-wing media ecosystem and played a prominent role in propagating Pizzagate and the Seth Rich conspiracy, responded to an interview question from the Evening Standard with “I love Russia Today!” The Evening Standard author described RT as “one of his most cited sources” and followed up with: “You are aware that it’s Russian propaganda, right?” to which Adl-Tabatabai responded, “Well, the BBC is British propaganda. It’s the same thing.”33 In other words, it would be a mistake to imagine that the primary role of RT is to propagate content directly through social media campaigns. Rather, the evidence suggests that RT intentionally serves as a source and hub of a much more diffuse network of sites that reuse narratives the station produces.

In the other direction, RT and Sputnik offer accreditation for others who carry the Russian line. For example, in the case of the campaign to smear the White Helmets, a humanitarian organization in Syria, RT and Sputnik gave significant prominence and repetition to blogger Vanessa Beeley, associate editor of 21stcenturywire.com, who was at the epicenter of the Russian-led disinformation campaign.34

(p.245) In our study of the “deep state” frame, we found that RT gave Ron Paul a platform to propagate the idea just after the election, that Sputnik published a piece by an editor of 21stcenturywire.com, and that both sites repeated this frame often throughout 2017. Zero Hedge offered guest publishing slots for The Saker (described in Chapter 5), who runs a site with quite a bit of Russian apologia, as well as an early piece from the Strategic Culture Foundation,35 a Moscow-based think tank that Politico described as a part of Russian influence efforts.36 Despite these traces of Russian efforts, the most important anchors of the deep state frame were American and came from diverse political orientations. The Virgil posts on Breitbart and its full-throated embrace by Fox offered a quite distinct frame from the Greenwald pieces. But it would be a stretch to imagine that either Greenwald or Breitbart needed Russian prompting to frame their criticism as they did. Similarly, in Chapter 7 we saw the central role that Cassandra Fairbanks, then working for Sputnik but publishing on WeAreChange, played in weaving together the spirit cooking and Pizzagate narratives. Reddit users and alt-right characters concocted those narratives from the Russian-hacked Podesta emails and fed them into stories that WikiLeaks, Infowars, and Drudge pushed up the propaganda pipeline.

From our observations of the last three years, there is one example that might count as a significant temporary success of the interplay between white, gray, and black propaganda. It involves the Guccifer 2.0–Adam Carter–Forensicator line of attack on the Russian origins of the DNC hack, which was interwoven with the Seth Rich conspiracy theory we encountered in Chapter 5, but was quite distinct. In particular this line of attack was more directed toward harnessing the national security skepticism that characterized the Greenwald posts than the partisanship of the Breitbart/Fox variety. We say that this story “might” count as a success because the evidence that it was primarily a Russian effort, as opposed to genuine pseudonymous critics of the national security establishment or fans of WikiLeaks, who were then aided at the margins by Russian propagandists, is circumstantial. We say the success was “temporary,” because it succeeded in changing the public narrative around the DNC leak for about two months in the middle of 2017 before the orientation of the narrative returned to the original configuration of the right in opposition to the rest of the media ecosystem.

As we described in discussing hacking and doxxing, by the time that the ODNI released its January 6, 2016, report, there was broad consensus that the DNC had been hacked by Russian operations and that the “Guccifer (p.246) 2.0” persona who claimed responsibility was a Russian cover (an allegation later elaborated and supported in a July 2018 indictment by the special counsel). On February 8, 2017, a person writing under the pseudonym Adam Carter first published what would become a regularly updated blog post describing the reasons to doubt that Guccifer 2.0 was in fact part of a Russian information operation. He argued that Guccifer 2.0 was an effort by the Clinton campaign or the DNC to point the finger at Russia and discredit WikiLeaks.37 The primary evidence was that metadata from the documents Guccifer 2.0 released suggested that the document was in fact created by a person with ties to the Democratic National Committee. On February 17, as the deep state narrative was spinning up in response to the Michael Flynn resignation, a different person going by the handle tvor_22 and the Cyrillic нет (nyet) published a later-deleted post on Medium that analyzed the documents made public by Guccifer 2.0. That post argued that the extremely obvious ways that Guccifer 2.0 “hid” clues of being Russian were best explained as efforts by the DNC to hide the origin of the leaks, by the CIA to besmirch Russia, or by Russia to sow disinformation and chaos.38 On March 8, tvor_22 published another deeply ironic post suggesting that the Russian Bear “paw prints” all over the alleged hacks were so laughably clumsy that they were more likely to be CIA false flag operations than actual Russian operations.39 Whether this account was a bemused hacker ironically taking on an obviously made-up Russian persona to state that the paw prints were too obviously Russian to have been really Russian, or whether it was a Russian information operation pretending to be such a bemused hacker, is anyone’s guess.

On March 20, 2017, the House Intelligence Committee ran its Hearing on Russian Active Measures. Representative Adam Schiff’s opening statement underscored the sustained Russian attack, connecting the DNC hack, the DCLeaks site, and the Guccifer 2.0 diversion operation as elements in the attack on the American presidential campaign.40 In response Sputnik News reporter Cassandra Fairbanks tweeted “Pretty sure Guccifer 2 is not a ‘creature of Russia,’ @RepAdamSchiff” and attached a screenshot of a direct message exchange she had had with Guccifer 2.0 on August 24, 2016. In that exchange she offered to interview Guccifer 2.0, and he answered that he would rather that she interview him for her WeAreChange blog, saying “I don’t like sputnik coz it’s Russian.”41 Who exactly might be persuaded by a Sputnik reporter sharing a purported screenshot from an alleged Russian agent claiming that he is not in fact a Russian agent, is a curious question. These Fairbanks tweets were then included in the Adam Carter blog as part of the evidence that Guccifer 2.0 was not a Russian job. The Senate Select (p.247) Committee on Intelligence then held its open hearings on disinformation and Russian active measures on March 30.

Following the congressional hearings, April saw an uptick in Seth Rich and Guccifer 2.0 conspiracies. On April 8, former Playboy model Robbin Young published a set of purported screenshots from Twitter direct message exchanges with Guccifer 2.0. In one of the messages Guccifer 2.0 asked for help investigating “the real story of his life and death . . . his name is seth, he was my whistleblower.”42 That same day the WikiLeaks Twitter account tweeted Young’s DMs, linking Adam Carter’s site, which hosted a copy of the DM exchange.43 Stories on Heat Street and republished on Zero Hedge with the clickbaity title ‘Guccifer 2.0’ Chat With Nude Model Sparks New Conspiracy Theories About Seth Rich Murder” followed,44 joining the Gateway Pundit.45 Fox and Friends then also reported on the Robbin Young–Guccifer exchange as raising the question of Seth Rich’s murder, although added that “it’s not clear if the messages are, in fact, authentic.”46 Toward the end of April, Cassandra Fairbanks returned to the conversation with two stories in Big League Politics. In the first, she used alleged screenshots from Twitter DMs with Guccifer. She described her relationship with Guccifer as involving the writing of a story about the leaks. “At this point,” Fairbanks wrote, “I was still supporting Sanders in the primary.” Among those chats, she reported, Guccifer 2.0 tied the death of Shawn Lucas, who had served papers on the DNC alleging primary fraud against Sanders, to that of Seth Rich.47 In the second, she reported on Adam Carter’s allegations that Guccifer had deliberately planted Russian fingerprints on his account.48 It was clear that at this point the Guccifer 2.0 story was trying to attract left-wing partisans whose skepticism of the national security establishment would outweigh their partisan proclivity to believe the worst of Donald Trump. Between the Guccifer 2.0 persona, Fairbanks’s position at Sputnik, the timing surrounding the congressional hearings into Russian interference, and the party most likely to benefit from this shift in blame, circumstantially this storyline exhibits more hints of a Russian effort than does the Seth Rich episode, which would soon took center stage in May and June.

On July 8, 2017, the New York Times published its first story on the famous Trump Tower meeting arranged by Donald Trump Jr. with Natalia Veselnitskaya.49 The next day, they followed up with reporting that Trump Jr. had been enticed to the meeting with the prospect of damaging information on Hillary Clinton. That same day, an individual working under the pseudonym the Forensicator published an analysis of the Guccifer 2.0 metadata, introducing a new factor into the analysis of the email leak/ (p.248) hack: the data transfer rate at which the files were copied was too high for any existing broadband connection to sustain, much less from an international connection.50 The story was then reported on Disobedient Media51—a website that is hard to peg as either clearly left or right—generally skeptical of powers that be, but with a heavy focus on WikiLeaks and Julian Assange (when we observed the site, four of twelve stories under “U.S. Politics” concerned Assange; another four covered the Nunes memo). Adam Carter is the site’s technology writer. Its cofounder and contributor, William Craddick, is also a contributor to Zero Hedge, where his profile proudly announces that he “[d]iscovered and published the Clinton-Silsby human trafficking scandal”52 that we encountered in Cassandra Fairbanks’s posts in Chapter 7.

From there, the story spread through the by now usual suspects on the right: the Free Republic, the Gateway Pundit, Zero Hedge, and BB4SP, with a Zero Hedge story getting the extra credit of citation by the Daily Stormer prefaced by “Well, if the Jews and their shills thought this was the end of their troubles, they were wrong.”53 Unsurprisingly, RT carried the story as well.54 The story also appeared on a subreddit for supporters of Bernie Sanders, r/WayOfTheBern. There, the post laid out the Adam Carter and Forensicator analyses and concluded: “In short, Russiagate is a hoax concocted by the DNC in collaboration with Crowdstrike, and given the stamp of approval by Deep State tools—hand-picked by James Clapper—eager to defame Russia. The Democrats and the MSM subsequently embellished this narrative by claiming that the Trump campaign somehow had ‘colluded’ in Russia’s non-existent interference.” (emphasis in original).55 The post was clearly trying to leverage the dual valence of the “deep state” frame—appealing to the older, more general anti-national-security-establishment frame—even as the frame continues to do work on the right in its new, more partisan form. Despite this one appearance on the left, the story remained almost exclusively in right-wing media for the next two weeks.

On July 24, 2017, a group of former U.S. intelligence professionals published an open letter to the president on consortiumnews.com, questioning the consensus view of the active intelligence community that the DNC hack was perpetrated by Russian intelligence.56 Veteran Intelligence Professionals for Sanity (VIPS) had formed to protest manipulation of intelligence that led to the Iraq War.57 The group who signed this open letter consisted, among others, of William Binney, Kirk Wiebe, and Ed Loomis, individuals who had been at the center of exposing the elements of the illegal surveillance programs initiated by the Bush administration after 9/11. These revelations later exploded in 2006, and parts were later corroborated in the (p.249) materials by Edward Snowden. These were no crackpots or Russian stooges but respected professionals with a personal history of calling foul, getting it right, and for some of them, being persecuted for it by the Bush and Obama administrations. Some were heroes to libertarians and civil libertarians alike. The memo relied on the work and logic of the Forensicator and Adam Carter, which identified the “key fact” that

someone working in the EDT time zone with a computer directly connected to the DNC server or DNC Local Area Network, copied 1,976 MegaBytes of data in 87 seconds onto an external storage device. That speed is much faster than what is physically possible with a hack.” (emphasis in the original). It thus appears that the purported “hack” of the DNC by Guccifer 2.0 (the self-proclaimed WikiLeaks source) was not a hack by Russia or anyone else but was rather a copy of DNC data onto an external storage device.

They surmised from the metadata that downloading the volume of materials in the specified time would have required a download speed of 22.7 megabytes per second (slightly over 180 megabits per second). Using average speeds from speedtest.net, the authors of the VIPS memo argued that such speeds were not available in the United States from broadband providers but that that speed was very close to the speed usually obtained in downloading to a USB thumb drive. The memorandum hypothesized from these facts that a likely scenario was that the DNC wove a story of Russian interference to discredit any wrongdoing that the emails exposed, as well as to turn it to the disadvantage of Donald Trump, tarring him with being the preferred candidate of Vladimir Putin. The story was initially republished on a few left-wing sites and on the Antiwar blog, got substantial coverage on the right, and was picked up by RT and Sputnik as well.

On August 9 The Nation published a long piece by Patrick Lawrence, detailing the VIPS claims as well as the underlying work that Forensicator and Adam Carter had published.58 The next day, Bloomberg published a column by Leonid Bershidsky. Bershidsky had been a journalist in Russia and wrote very publicly about his disillusionment with Russia when he emigrated to Germany after the Crimea annexation. His column suggested that the VIPS analysis and The Nation story demanded more serious coverage from traditional media. He further argued that the participation of such highly respected professionals warranted more than the back-of-the-hand treatment they had received from mainstream media.59New York Magazine and the (p.250) Washington Monthly published stories criticizing The Nation story, but with little analysis.60 Salon published a piece summarizing the VIPS argument, claiming that mainstream media were ignoring the story because it was an embarrassment that they had all rushed to the wrong judgment and lamenting that the VIPS memo was receiving attention almost exclusively by right-wing media.61

Indeed, on the right all the top sites immediately picked up The Nation story and used its provenance as a major source of legitimation as a Breitbart headline makes clear: “Left-Wing Magazine The Nation Report Puts ‘Russian Hack’ DNC Narrative in Freefall.”62 Tucker Carlson interviewed Bill Binney on his show and framed the VIPS claims as “not just that the President Trump didn’t collude with Russia, you’re calling into question the core allegation; that Russia is responsible for hacking into the DNC server.”63 The VIPS memo did nothing of the sort. It made no claims at all about the broader “collusion” question and focused very specifically on evidence regarding the DNC hack. Nonetheless, the chyron filling the bottom of the screen cycled between “Former NSA Technical Director, Trump-Russia Collusion Story is Bogus, Agenda-Driven,” and all caps “FMR NSA TECH DIR: NO PROOF OF COLLUSION.”

Are the Russians Coming?

Figure 8.1 Bill Binney interview on Tucker Carlson, August 14, 2017.

Figure 8.1 bears attention. Carlson hosted Binney for five and a half minutes. For much of this time Binney focused narrowly and technically on the forensic evidence surrounding the DNC hack itself. His jargon was likely opaque to most TV viewing audiences. Carlson nodded gravely and (p.251) listened intently. Binney said nothing about the Trump Tower meeting or the question of whether or not there was or was not collusion between the Trump campaign and Russian state operatives. But the core message was impossible to miss. An unimpeachable expert offered detailed technical analysis that refuted the core claim of the Trump-Russia investigation.

Figure 8.2 describes the link network of stories surrounding the Guccifer 2.0 analysis that resulted in this remarkable interview. It leaves no doubt as to the centrality of the interventions by Forensicator (both the WordPress blog and the publication on Disobedient Media) and by the Adam Carter blog (g-2space). It leaves no doubt that The Nation adoption of this storyline was key. The Intercept, as we will soon see, was actually part of the rejection of the theory, not its adoption, and came much later, in the autumn of 2017.

Are the Russians Coming?

Figure 8.2 Network map of stories that mention the Forensicator theory of Guccifer 2.0. Architecture and node sizing by media inlinks.

In the meantime, it seems that some of The Nation’s staff were rebelling against the publication of Lawrence’s piece. By August 15, the Washington (p.252) Post reported that, under internal pressure from journalists at The Nation, the magazine was conducting a post-publication review.64 On September 1, 2017, The Nation appended an editor’s note to the original August 9 story and published a collection of pieces on the VIPS memo.65 Those pieces included the results of an independent investigation of the VIPS memo that The Nation had asked Nathan Freitas of the Guardian Project to conduct, a countermemo by dissenting members of VIPS, including some like Thomas Drake and Jesselyn Radack, who had long been allied with Binney and other signatories on the original VIPS memo, and a response by the authors of the original VIPS memo. The Freitas memo offered a detailed technical analysis that explained how the Forensicator memorandum went well beyond the data available and how the original VIPS memo had extrapolated further beyond the Forensicator’s analysis.66 The VIPS dissenters also focused on the extent to which the Binney group had jumped to conclusions beyond those warranted by the Forensicator’s analysis, rather than raising the less sensational but legitimate questions of why the intelligence community has not released detailed evidence supporting the claim that Russians had hacked the DNC servers.67 The original VIPS memo authors responded, primarily emphasizing the devastating damage that the intelligence errors leading up to the Iraq War had caused and warning of the danger that we were, once again, suffering from a similar bout of groupthink.68

Major media outlets reported on the follow-up coverage in The Nation, which was viewed as a near-retraction of the original story in The Nation. Right-wing media covered the same collection of stories as an indication that The Nation caved in to mainstream anti-Trump media pressure. The story re-emerged in November, when two veteran national security reporters, Duncan Campbell and James Risen, reported in the Intercept that CIA Director Mike Pompeo had met Bill Binney at President Trump’s request. The most notable feature of this story was that it was published in the Intercept, founded by Glenn Greenwald, who had been the most important source of skepticism of the Russia hypothesis outside of the right-wing media ecosystem, and who had long worked with national security whistle-blowers among the VIPS memo authors and dissenters. The November 7 story about the Binney-Pompeo meeting, however, emphasized the extent to which the VIPS-Binney-Forensicator theory had by now been abandoned and used the meetings primarily as a signal of concern that Pompeo was bending the CIA to serve Trump’s agenda rather than preserving the independence of the agency. Other publications in the mainstream and left emphasized the impropriety (p.253) of the Trump’s intervention and the danger posed by a CIA director willing to bend the agency’s mission to help the president politically. Publications on the right, by contrast, framed the meeting as further support for Binney’s theories.

The Guccifer 2.0 case and those of Infowars, 21stcenturywire.com, True Pundit, YourNewsWire.com, and Zero Hedge, all present the basic difficulty of sorting out which of these are gray or black propaganda outlets, as opposed to honest critics promoted by propagandist outlets on one hand, and useful idiots just out to make a point or a buck, on the other.

True Pundit, a site founded in March 2016, has grown explosively during the first year of Trump’s presidency and was identified by the German Marshall Fund’s Hamilton 68 project (an online dashboard offering visualizations of “topics and URLs promoted by Russia-linked influence networks on Twitter”) as one of the sites most widely promoted by Russian social media operations.69 It showed up in our observations as one of the most tweeted sites involved in the propagation of the deep state frame. Practically nothing is known about the origins or authors of this successful site. A bit more is known of YourNewsWire.com. The site was described by Snopes as a “notorious fake news generating site,” in the context of refuting its theory that the June 2017 suicide bombing at the Arian Grande concert in the Manchester Arena, in which 22 people were killed and over one hundred injured, was a false flag operation.70 It has the dubious honor of a perfect “pants on fire” score from PolitiFact.71 A Hollywood Reporter profile of the two owner-authors, a couple living in Los Angeles, noted in passing that an unnamed British paper had reported that “a European Union task force set up to combat Russian propaganda had classified the outlet as a proxy.”72 Digging deeper into the British press coverage, however, about the Brit who moved to Los Angeles with his American husband and makes a living by spinning checkout-counter-tabloid material, some of which is written by his mother, suggest that Sean Adl-Tabatabai is more an Alex Jones wannabe than a chapter from The Americans.73 Because of all the negative coverage, Google removed the site from its advertising network in 2017, and it remains to be seen what impact this will have on the site. While the couple told Hollywood Reporter that the move had cost them a 60 percent drop in their revenues, data from the first quarter of 2018 show that the site had retained its level of traffic, with monthly visits roughly halfway between the Weekly Standard and Reason magazines—a position it had held without much variation throughout the preceding 18 months. One of Zero Hedge’s lead authors (p.254) resigned from the site in mid-2016, telling Bloomberg in the process that he was fed up with cranking out stories that fit the model of “Russia=good. Obama=idiot. Bashar al-Assad=benevolent leader. John Kerry=dunce. Vladimir Putin=greatest leader in the history of statecraft.”74 Nonetheless Lokey, the former Zero Hedge contributor, also told Bloomberg that these guidelines were honed carefully to increase clicks rather than for a political purposes.

Perhaps investigative journalists or law enforcement agencies can find smoking gun evidence that one or another of these actors is in fact a Russian agent. We have not found such evidence in the public record. But the actual observed patterns of behavior suggest that whether there is a formal relationship or whether these sites merely repeat the party line because it fits their ideology or contributes to their bottom line is substantially less important than understanding that these sites behave in this manner and that we are able to tag such sites as repeat offenders when it comes to spreading Russian propaganda. For purposes of legal responsibility, knowledge and intent are critical. For purposes of understanding what sites often serve as pathways for Russian propaganda, witting, unwitting, or witless matters a great deal less than the fact of being such a conduit.

Caution, After All

We began this chapter with the claim that one would have to be willfully blind or complicit to deny the fact that there has been a sustained Russian attack on the American media ecosystem aimed at sowing division and disinformation. The evidence we described in condensed form should be enough to support that assertion. There are simply too many diverse sources to ignore: government and nongovernment, for profit and nonprofit, defense-centered and internet-research-centered, American and European, all of which confirm the general pattern and many of the details that support the claim that Russia mounted a sustained campaign against the United States as well as various countries in Europe.

But evidence of sustained effort is not the same as evidence of impact or prevalence. It would be profoundly counterproductive to embrace the narrative that we can no longer know what is true because of Russian bots, sockpuppets, or shady propaganda. Indeed, having us adopt that attitude would mark a remarkable success for that Russian effort: the success in denying democracy one of its core pillars—the capacity to have a public debate based (p.255) on some sense of a shared reality and trust in institutions. To be clear, we do not claim that the following pages “prove” that Russian efforts didn’t matter. We offer them rather as examples of the kind of perspective and scrutiny we think necessary before one moves from alert to panic in understanding the threat of the attack.

Hacking and Doxxing Minimal Effects

While the Podesta emails may have been released primarily to rally the base, it is very clear from the timing and context that the DNC email hack and its release on DCLeaks and WikiLeaks was intended to deny Clinton the support of disappointed Sanders supporters. In that regard, it represented a relatively obvious target, given the contentiousness of the primary, and, as we saw in Chapter 6, paralleled the strategy Breitbart pursued with the release of the Clinton Cash video that same weekend before the convention.

How important were these hacking attacks? We have already seen in Chapter 6 that “emails” was the word most closely associated with Clinton in the run-up to the election, but we also saw that media attention was primarily driven by State Department releases and to a lesser extent Judicial Watch Freedom of Information Act (FOIA) victories. The DNC and Podesta emails certainly produced increased attention, particularly the latter, but the response even to the Podesta emails was dwarfed by the Comey announcement about reopening the State Department server investigation. The emails certainly provided fodder for right-wing media. They were at the heart of Pizzagate. But they were not at all at the root of the much more widely held Clinton-pedophilia conspiracy theory concerning the Lolita Express, which, as we saw in Chapter 3, was rooted in FOIA-enabled reporting by Fox News using flight logs and earlier Gawker reporting. The Seth Rich and Guccifer 2.0 conspiracies certainly played repeated roles over the course of 2017 as means of diverting attention from Russia-related news that was embarrassing to Donald Trump, but, as we saw in Chapter 5, the heavy lifting on Seth Rich was done by Fox rather than by Russian active measures.

If we look at the direct functional target—splitting Sanders supporters from Clinton—the hacking operations seem to have failed. Reviewing the coverage of the DNC email leaks reveals that, as we have repeatedly seen, the media ecosystem occupied by the Sanders left lacked the kind of conspiracy-amplification sites so typical of right-wing media. The stories from the emails did in fact emerge and were covered by both left and center-left media. But (p.256) then they receded. The most widely shared email from the DNC dump was an email from Amy Dacey to Brad Marshall, both of the DNC, entitled “No shit.” In that email Marshall proposed planting a question in a Sanders event in which Sanders would be asked about his religion. Marshall wrote, “I think I read he is an atheist. This could make several points difference with my peeps. My Southern Baptist peeps would draw a big difference between a Jew and an atheist.” This original email was covered by the Intercept, Washington Post, CNN, and Politico, coverage that was quite prominent when measured by media inlinks (Figure 8.3a). Marshall then followed up with “It’s the Jesus thing,” to which Dacey responded, “AMEN.”75 This latter part of the exchange got more attention on Facebook generally, in particular on the right (Figure 8.3b).

Are the Russians Coming?

Figure 8.3a Media sources that linked to the initial leaked “No shit” email. Nodes sized by number of media inlinks.

Are the Russians Coming?

Figure 8.3b Media sources that linked to the final “AMEN” response in the exchange. Nodes sized by number of Facebook shares.

After the mainstream coverage, however, no major Facebook- or Twitter-driven sites kept the story alive on the left, so the impact of the story was small. We see an even starker rightward bias when we look at the other major (p.257) Facebook-shared email that could have inflamed the left. Entitled “re: Alaska ‘Counter’ event,” the email mentioned an informant inside the Bernie campaign helping the DNC. Here coverage stayed purely within the right-wing media ecosystem, primarily the Gateway Pundit and to a lesser extent sites like beforeitsnews.com and dcclothesline.com (Figure 8.4).

Are the Russians Coming?

Figure 8.4 Media sources that linked to the leaked “Re: Alaska ‘Counter’ Event” email. Nodes sized by number of media inlinks.

To the extent that the DNC leaks were intended to splinter the Democratic Party, they appear to have largely failed, and they appear to have failed largely because the media ecosystem on the left did not follow the same practices that were available for the propagandists to harness when they aimed at the right. The stories were reported in major publications and minor; they took their course; but they did not devolve into polarizing hatred-inducing stories of the kind we saw in Chapters 3, 4, and 7.

(p.258) Infiltrating Social Media

Using Social Media to Introduce Frames

Claims about social media amplification need to address not only the presence of a campaign but also the impact that a campaign has on framing, agenda setting, or behavior on a scale large enough to shape events. We can take, for example, one of the very specific allegations in the Mueller indictment: “Defendants and their co-conspirators began purchasing advertisements that promoted a post on the ORGANIZATION-controlled Facebook account ‘Stop A.l.’ The post alleged that ‘Hillary Clinton has already committed voter fraud during the Democrat Iowa Caucus.’ ” The (p.259) indictment went on to identify communications by the Russian-controlled Twitter handle @Ten_GOP on August 11, and then again on November 2, alleging primary and intended general election voter fraud by the Clinton campaign.

For purposes of understanding whether the acts are illegal, the marginal impact on agenda setting during an election cycle is irrelevant. If an accountant steals $5,000 from a billionaire client, her theft is no less a crime for the fact that the billionaire will not miss the $5,000. But if we are trying to put the acts in context, we need to understand what role that particular intervention might have played in the overall “voter fraud” frame and how it played out in the election. Looking at our election data, Figure 8.5 makes clear that there is in fact a distinct uptick in the use of the “voter fraud” in early August (you need to look carefully not to be confounded by the huge upsurge in mid-October).

Are the Russians Coming?

Figure 8.5 Sentences per day mentioning “voter fraud” during the election period.

If Russians injected that meme into the American election, even if they were not responsible for the later upsurge, that would indeed be a major concern. Zooming in on mid-July to mid-August, however, Figure 8.6 suggests that the uptick in use of the term “voter fraud” precedes August 4, when the indictment alleges a Russian Twitter campaign to push the term. Indeed, the term peaks on that day and declines thereafter.

Are the Russians Coming?

Figure 8.6 Sentences per day mentioning “voter fraud,” July 15–September 20, 2016.

What precipitated that spike in early August was not Russian bots, sockpuppets, or targeted advertisements but a candidate’s strategy. On July 27, during a Reddit AMA (Ask Me Anything) forum, Trump was asked, “We firmly believe Hillary will try and steal this election through vote fraud, especially given recent events. What is your campaign doing to ensure that we have a fair election?” His answer was, “Voter fraud is always a serious concern (p.260) and authorities must be vigilant from keeping those from voting that are not authorized to do so.”76 The candidate’s assertion of voter fraud got coverage in the Washington Post, on ABC News, on the “PBS NewsHour,” in the Guardian, and in the Miami Herald, alongside the New York Post.

On July 29 a court in North Carolina struck down that state’s voter ID law, holding that it had discriminatory intent and targeted African Americans “with almost surgical precision.”77 That same day courts in Wisconsin and Kansas similarly struck down as unconstitutional laws passed by Republican state legislatures involving voter ID and other restrictions. These decisions were widely reported across the media and launched a series of attacks by the Trump campaign. That same day, Breitbart published an interview with Roger Stone under the title “Roger Stone on the Milo Show: How Trump Can Fight Voter Fraud.”78 There Stone alleged that voter fraud was widespread and stated that, “[i]f there’s voter fraud, this election will be illegitimate, the election of the winner will be illegitimate, we will have a constitutional crisis, widespread civil disobedience, and the government will no longer be the government.”79

At a campaign rally on August 1 Trump picked up the theme, alleging that the Democratic primary had been rigged and that “I’m afraid the election is going to be rigged, I have to be honest.”80 Later that day he appeared on Hannity81 and claimed that the 2012 election and Democratic primary had been rigged and that the general election might be. The next day Trump appeared on the “O’Reilly Factor,”82 then still the most popular show on Fox News,83 claiming that people were going “to vote 10 times” because of the court decisions striking down the voter suppression laws in North Carolina, Wisconsin, and Kansas. O’Reilly and Hannity each boasted audiences of roughly three million viewers. Rush Limbaugh ran a segment that day (p.261) that repeated Trump’s campaign speech statement. Although Limbaugh, not a Trump supporter in the primaries, questioned the campaign tactic as defeatist, he nonetheless reinforced the message by saying of Trump’s claim, “And, of course, there’s gonna be some people who believe this because there have been elections where there has been fraud. I mean, everybody knows this.”84 Limbaugh’s audience, though hard to gauge accurately, is estimated to be between 15 and 20 million listeners a week. On that same day, Trump again repeated the claim in an interview published in the Washington Post, saying, “there’s a lot of dirty pool played at the election, meaning the election is rigged.”85 The following two days, coverage of this voter fraud claim in its various iterations spiked across the entire political spectrum. The proposition that Russian ads on a Facebook account had a meaningful impact on pushing the voter fraud frame after it had been pushed repeatedly by the Republican candidate and covered across the major media ecosystem for several days, including the most highly watched right-wing talk shows, strains credulity.

The same goes for the alleged November 2 tweet that the indictment describes. The tweet from @TEN_GOP stated “#VoterFraud by counting tens of thousands of ineligible mail in Hillary votes being reported in Broward County, Florida.” But again, the big spike in coverage of voter fraud followed an early morning October 17 tweet, in which Trump wrote: “Of course there is large scale voter fraud happening on and before Election Day. Why do Republican leaders deny what is going on? So naive!” Nancy Pelosi and Harry Reid released a joint statement in response,86 calling on Republicans to affirm the fairness of the democratic process. Broad public coverage ensued. By November 2, the issue had mostly receded yet again, and again, it is difficult to conclude that the @TEN_GOP had a significant incremental contribution. Moreover, Pelosi and Reid were fighting an old battle that had simply been rekindled by Trump in his various comments, both in August and in October. The North Carolina, Wisconsin, and Kansas laws which were struck down were all based on the oft-repeated but never proven claim that voter fraud is widespread in the United States. An NBC story by Zachary Roth that sought to put Trump’s August 1 and 2 comments in context,87 emphasized that it recapitulated misstatements that John McCain made in 2008.88 Those misstatements had resulted in a Public Policy Polling finding in 2009 that over 50 percent of Republicans believed Acorn had committed voter fraud and stolen the election for Obama.89 He traced the claims further back yet, to the 2004 Bush campaign and the 1996 Bob Dole campaign as well. The Russians pushed a false narrative persistently repeated by Republican elites (p.262) in an effort to justify laws known to disproportionately decrease eligible Democratic votes, particularly among Black and Latino populations. They introduced nothing new. As Limbaugh said, “I mean, everybody knows this.”

Using Social Media to Suppress and Split the Democratic Vote

The indictment also alleges that the Russians used Instagram accounts they called “Woke Black” and “Blacktivist” to send messages to dissuade black voters from voting or to divert their vote to Jill Stein. If we consider Jill Stein, in the three states where her vote margin was larger than the Trump/Clinton difference—Michigan, Wisconsin, and Pennsylvania—at least according to exit polls, Stein drew 1 percent or at most 2 percent of the Black vote, less than pre-election polling suggested.90 More importantly, a pre-election Bloomberg report on the Trump campaign’s media campaign quoted an unnamed “senior official” as saying, “we have three major voter suppression operations underway,” one of which was to emphasize Clinton’s use of the term “superpredator” in 1996 to discourage black voters from showing up at the polls.91 That report described in detail a campaign, using Facebook to post messages and videos that featured the recording of Clinton’s comment, targeted using Facebook’s capabilities to make sure that “only the people we want to see it, see it.” The interviewee from the campaign boasted that “[w]e know because we’ve modeled this” that “[i]t will dramatically affect her ability to turn these people out.” And that same report quotes Steve Bannon, then running the Trump campaign, as saying, “I wouldn’t have come aboard, even for Trump, if I hadn’t known they were building this massive Facebook and data engine.” Again, as with the “voter suppression” frame, it is difficult to lend credence to the proposition that the Internet Research Agency, through Instagram posts, made a meaningful marginal impact in comparison to the directed efforts of the Trump campaign working with Facebook’s political marketing team.92

Unlike many other aspects of the allegations, however, the questions of voter suppression through targeted advertising are eminently resolvable, if we have the right disclosure system in place. Unlike Twitter, Facebook actually has the real identity of every account holder. And Facebook also has a record of the entire set of communications to which an account holder was exposed. It would not be too difficult to run a comparison between all the accounts exposed to Russian advertising interventions and accounts of otherwise demographically similar matched accounts—say, Black registered voters in Florida in the same congressional districts or smaller geographic settings—and then match the voter rolls with the real identities of the Facebook account (p.263) holders. We could then know, to a high degree of confidence, whether being exposed to these Russian ads was significantly associated with lower turnout than among matched Facebook users who were not exposed to the advertising. It is precisely to conduct such investigations that it would be important to produce a record of all targeted advertising, and to make it available, under constraints appropriate to preserve the privacy of the individuals and the integrity of the voting process, for researchers who study the effect of online communications on American democracy.

Real-World Protests

Perhaps the most exotic Russian interventions involve the orchestration of real-world rallies. These are included in the Mueller indictment, in the congressional hearings record, and in several news reports. The idea that Russians impersonating Americans get real Americans to actually get out on the streets and protest has a spy movie quality to it that is irresistible. Two cautions must, however, be kept in mind. First, the reported demonstrations were a drop in the bucket on the background of the steady stream of campaign events, not least among these mass appearances by the candidate or surrogates in which masses of people shout in unison, “Lock her up! Lock her up!” At the time, the Russia-induced protests drew little attention, and it is hard to imagine that this handful of secondary pro-Trump or anti-Clinton events had a meaningful impact on the election or the level of animosity between the two parties in the run-up to the election. The second is that these protests were not only relatively small, but whatever traction they had was in the context of a much broader and deeper political animosity that was not of Russian making.

The very first such report, in the Daily Beast, concerned an effort by the Secured Borders persona to mount an anti-immigration rally in Twin Falls, Idaho.93 As a Daily Beast story explained, Twin Falls had been the target of a long-standing hate campaign starting with a January 2016 story that reported on Chobani’s plan to hire immigrants, entitled “American Yogurt Tycoon Vows to Choke U.S. With Muslims.” Infowars ran stories, among them pearls such as “500% increase in tuberculosis in Twin Falls,” and “Idaho Yogurt Maker Caught Importing Migrant Rapists.” Chobani sued, and as with Pizzagate, Alex Jones was forced to settle.94 Breitbart, for its part, fanned the flames of anti-immigrant fervor with a different hyped-up version of a case of sexual assault dealt with by Idaho’s juvenile justice system.95 Despite the fertile ground laid by right-wing media, only 48 people registered at the event site, and, according to the Daily Beast, only four claimed to have gone. The (p.264) city’s mayor, however, told the New York Times that “after two years of ‘robust debate’ over the city’s refugee resettlement program, which dates to the 1980s, it was ‘kind of surreal’ to discover that Russia had joined in.”96 Again, though, it’s important to recognize that however surreal the discovery, its actual impact on the debate in Idaho or in Twin Falls was, for all practical purposes, nil.

Twitter Bots and Shaping the Debate

Finding bots is hard. Finding Russian bots is even harder. They operate on a very large scale and therefore are only amenable to automated searching. In order to train a machine learning algorithm, one needs a large training set of accounts one knows for sure are “bots” or “Russians” and ultimately both. This is the “ground truth” that is the weakest link—because no matter how sophisticated the algorithm, or how many data points it relies on, all the algorithm does is identify more instances of the entities that were in its training set. If we are uncertain as to whether the entities in the training set are bots or Russians, we will be equally uncertain about the entities the algorithm identifies as similar to them. There are several efforts aimed at identifying such accounts. Some simply look at the posting rate of Twitter accounts and assume that accounts tweeting more than some threshold number of posts are bots.97 We used a variant of this approach to filter the most obvious bots from our Twitter maps. Some, most prominently the Indiana University group headed by Filippo Menczer that developed the Botometer and Hoaxy, and its graduates, try to diagnose individual accounts based on a broad range of characteristics.98 Again, we used a variant of this approach as well to test whether bots influenced our overall structure. In the Russian context, efforts by Joshua Tucker’s lab at New York University (NYU) have provided some powerful initial results.99 Others try to take an approach that moves away from diagnosing whether an individual account is a bot or not, and instead asks whether a network of accounts is a coordinated effort, whether fully automated or not.100 Similarly, distinguishing whether an account is or is not “Russian” as opposed to, say, a party whose interests are aligned with those of Russian propagandists is at least as difficult as deciding whether a clickbait factory is a Russian propaganda site or merely a nihilist out to make a buck. The Hamilton 68 project, which has garnered a lot of media attention because of its attractive visualizations and ease of use, relies on 600 hand-selected accounts whose nature and justification the project will not disclose publicly. They may be completely right, but it is difficult to rely on such an opaque process that purports to rely on such a small set of accounts. By contrast, the (p.265) SMaPP Lab project at NYU analyzes accounts that announce their primary language as Russian, and so is very likely to be right about the Russian origin of the accounts, and uses a robust hand-coding approach in determining what is, and is not, a bot; but as its creators clearly explain, this will not cover most of the interventions in the U.S. context, which are masked as Americans. Several researchers take an approach similar to the one taken by Young Mie Kim and her colleagues to quantify Russian Facebook advertising, which we noted earlier in this chapter—that is, using as their basis Facebook or Twitter accounts identified publicly by the companies or congressional committees as “Russian.” This approach is reasonably sound to set a lower bound on the activity but will clearly understate its actual level.

The projects we describe are the best of class in what they do at present. They should be encouraged and improved, published and cited. But they should not form the basis of reporting, with the kind of certitude one often reads in media accounts, that some large percentage of Twitter activity or a particular campaign are bots at all, much less Russian bots. The space is in continuous flux; the campaign strategies of the propagandists shift and mutate, and the research is getting better fast. For all we know, by the time our book is out the problem of bot detection and Russian operations detection will be solved. But we doubt it.

Beyond identification, the more important question is effect. The support of an army of bots does not necessary mean that a story is going to take off and be picked up by real accounts. For example, we analyzed all the emails on WikiLeaks, including those released by the State Department associated with the Clinton private server, the DNC hack, and the Podesta hack, and all the stories that linked back to any one of those emails, to create a bimodal map of emails and stories. This map allows us to identify the emails that appeared most frequently in media stories, who had produced the story, and the extent to which those stories had influence by any of the measures we have been using throughout the book—Facebook shares, tweets, and media inlinks. One of the most curious standouts is an email included in only six stories. By comparison, the most covered emails appear in anywhere from 60 to over 240 stories.

Yet one of the stories about this email was among the most highly tweeted stories in our entire set. The email was a 2010 email, related to the Chelsea Manning leaks to WikiLeaks. It was dated five days before the publication of the embassy cables leak, but after publication of the Afghanistan and Iraq war logs leaks, from then-Director of Policy Planning at the State Department, to then-Secretary of State Clinton, attaching a memo outlining “possible legal and nonlegal strategies re wikileaks.”101 The story that was most promoted on (p.266) Twitter was from True Pundit, emphasizing that Clinton was willing to use “nonlegal means” as evidence for the core claim in the story—that Clinton was promoting a drone strike on Julian Assange (then in London . . .) to stop publication of the embassy cables.102 Given that the claim that Clinton wanted to order a drone strike on Assange got no traction more broadly, this particular case suggests some caution before assuming that a substantial Twitter promotion campaign will necessarily promote a story in the media agenda. The sparse attention from other media stories and the minuscule number of stories about it (Figure 8.7), combined with the remarkably high tweet count strongly suggest that the story was the subject of a synthetic Twitter promotion campaign. While it is possible that such a campaign could (p.267) reflect True Pundit’s purchase of a viral marketing campaign, none of the other stories True Pundit published about the emails received more than one or two tweets. More likely, given that RT was the only other site to report on this email, is that the Twitter campaign was in fact Russia-directed. This would also be consistent with a later 2017 finding by Hamilton 68 of a brief period when True Pundit was again highly pushed by what they define as Russian bots.103 And again, given that no other True Pundit story in our email set was accompanied by a similarly large Twitter campaign, it seems more likely that Russian propagandists found a story that they decided to promote, but that it turned out to be of little interest to the right-wing media sphere as a whole, apart from other Russian propagandists.

Are the Russians Coming?

Figure 8.7 Map of emails hosted on WikiLeaks and sources linking to them.

The only two sites that reported on the “legal and nonlegal means” email from the State Department–released set were True Pundit and RT; the True Pundit story was among the most tweeted among stories that linked to an email.

The basic point is a simple one. We are not denying that there have been repeated Russian Twitter and Facebook campaigns aimed to push one narrative or another. It is worthwhile to try to identify these and to flag them as Russian efforts, so as to enable Americans who want to avoid being drawn in, because of the Russian origin of the attempt, from being duped. But it is critical to understand that without pickup from more influential media somewhere in the American public sphere, these Russian efforts will languish unnoticed. In order for Russians to be influential, their efforts must flow through the media ecosystem Americans inhabit. It is there that stories flourish or falter. The American media ecosystem is as resilient to Russian propaganda as it is resilient to all other falsehoods, whatever their source. And as we saw in our discussion of Russian white and gray propaganda, there is a part of the American media ecosystem for which knowing that the origin of a story is Russia is no reason to ignore it, as long as it aligns with the tribal narrative they are pushing. Willing embrace of divisive Russian propaganda, not innocent error because of Twitter and Facebook manipulation, is the core challenge.

Looking at the Russian interference in the U.S. 2016 election suggests three conclusions. First, both Americans and Europeans assessing the Russian threat need to be cautious in assessing the actual danger. Many current efforts are, justifiably, focused on the genuinely challenging problem of detection. Doing this critically important work creates a strong bias to assume that the hard-won successful observations of intervention are a sign of large impact and threat. However, trying hard, as the Russians clearly are, does not equal actual success in affecting the outcomes or attitudes of a society at large. And other countries, particularly those that have a pattern of trust in major media more similar to Germany’s, like Sweden or Canada, may be a lot more resilient to information operations than countries with more dysfunctional media (p.268) ecosystems, like the United States. Each country has to make its own internal assessment of its genuine susceptibility and tailor its responses to its real threat model, rather than to the perceived technical possibility of intervention as itself the threat. Second, given that generalized trust is an important ingredient in any democratic society, it is important not to overstate the impact of Russian propaganda and thus feed that generalized distrust. If Russian propaganda is in fact shaping events, we need urgent action. If instead it is simply an ever-present irritant, we need to put it in its right place. Inducing overreaction would be a major success for Russian information operations.

But the most important implication is for American conservatives, particularly economic and national security conservatives rather than white-identity conservatives. They have a genuine conflict on their hands. There is mounting evidence that “the Fox News effect” has given the Republican Party a clear edge in the past several election cycles. There is, it seems, a clear short-term partisan advantage of going along with the style and focus of these right-wing media outlets. But competition among outlets seeking to attract conservative audiences has resulted in a feedback cycle, as sites vie to produce more outrage and anger and get ever more extreme in their framing. If Breitbart’s third-most shared headline on immigration is “Six diseases return to U.S. as Migration Advocates Celebrate ‘World Refugee Day,’ ” then Gateway Pundit will respond with its second-most shared headline on the subject, “Obama Changes Law: Allows Immigrants with Blistering STDs and Leprosy into U.S.” And so the cycle continues. Infowars may circulate conspiracy theories about “Jeb Bush: Close Nazi Ties Exposed,” but it is Fox News hosts who consistently and repeatedly imply that the investigation into Russian interference is a “deep state” conspiracy to reverse the outcome of the 2016 election. This competitive dynamic among right-wing media increases the shrill, conspiracy-tainted tone and content of coverage and makes right-wing audiences ever more susceptible to manipulation.

The result is a United States that is vulnerable to disinformation campaigns, both foreign and domestic. That susceptibility does not come from Russia, though Russia clearly has been trying to exploit it. That susceptibility does not come from Facebook, though Facebook has clearly been a primary vector online. It comes from three decades of divergent media practices and consumption habits that have left a large number of Americans, overwhelmingly on the right of the political spectrum, vulnerable to disinformation and ready to believe the worst, as long as it lines up with their partisan identity. And that susceptibility should be, in the long term, unacceptable to conservatives every bit as much as it is to all other Americans despite its short-term electoral benefits.

Notes:

(1.) Peter Pomerantsev, Nothing Is True and Everything Is Possible: The Surreal Heart of the New Russia (United States of America: PublicAffairs, 2014).

(2.) Ben Nimmo, “Anatomy of an Info-War: How Russia’s Propaganda Machine Works, and How to Counter It,” May 19, 2015, http://www.stopfake.org/en/anatomy-of-an-info-war-how-russia-s-propagandamachine-works-and-how-to-counter-it/, quoted in Keir Giles, Handbook of Russian Information Warfare, NDC Fellowship Monograph Series 9 (NATO Defense College, Research Division, 2016), https://krypt3ia.files.wordpress.com/2016/12/fm_9.pdf.

(3.) Bruce Etling et al., “Public Discourse in the Russian Blogosphere: Mapping RuNet Politics and Mobilization,” Berkman Center Research Publication (Berkman Klein Center for Internet & Society at Harvard University, October 19, 2010), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1698344.

(4.) Darrell M. West, “President Dmitry Medvedev: Russia’s Blogger-in-Chief,” Brookings (blog), November 30, 2001, https://www.brookings.edu/opinions/president-dmitry-medvedev-russias-blogger-in-chief/.

(5.) Vladimir Barash and John Kelly, “Salience vs. Commitment: Dynamics of Political Hashtags in Russian Twitter,” Berkman Center Research Publication (Berkman Klein Center for Internet & Society at Harvard University, April 10, 2012); John Kelly et al., “Mapping Russian Twitter,” Berkman Center Research Publication (Berkman Klein Center for Internet & Society at Harvard University, March 23, 2012), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2028158; Sergey Sanovich, “Computational Propaganda in Russia: The Origins of Digital (p.417) Misinformation,” Computational Propaganda Research Project (Oxford Internet Institute, June 19, 2017), http://comprop.oii.ox.ac.uk/wp-content/uploads/sites/89/2017/06/Comprop-Russia.pdf.

(6.) @DFRLab, “The Russians Who Exposed Russia’s Trolls,” DFRLab (blog), March 8, 2018, https://medium.com/dfrlab/the-russians-who-exposed-russias-trolls-72db132e3cd1.

(7.) Ronald Deibert and Rafal Rohozinski, “Beyond Denial: Introducing Next-Generation Information Access Controls,” in Access Controlled: The Shaping of Power, Rights, and Rule in Cyberspace, ed. John Palfrey et al. (Cambridge: The MIT Press, 2010). Anna Polyanskaya, Andrei Krivov, and Ivan Lomko, “Commissars of the Internet: The FSB at the Computer” (Vestnik Online, April 30, 2003), http://www.vestnik.com/issues/2003/0430/win/polyanskaya_krivov_lomko.htm.

(8.) Id., at 28.

(9.) Evgeny Morozov, The Net Delusion: The Dark Side of Internet Freedom (New York: PublicAffairs, 2011).

(10.) Keir Giles, “HANDBOOK OF RUSSIAN INFORMATION WARFARE,” n.d., 6–7; Timothy Thomas, “The Evolving Nature of Russia’s Way of War,” Military Review, no. July–August 2017 (2017): 34–42.

(11.) Michael D. Shear and Matthew Rosenberg, “Released Emails Suggest the D.N.C. Derided the Sanders Campaign,” New York Times, July 22, 2016, sec. Politics, https://www.nytimes.com/2016/07/23/us/politics/dnc-emails-sanders-clinton.html; Jonathan Martin and Alan Rappeport, “Debbie Wasserman Schultz to Resign D.N.C. Post,” New York Times, July 24, 2016, sec. Politics, https://www.nytimes.com/2016/07/25/us/politics/debbie-wasserman-schultz-dnc-wikileaks-emails.html.

(12.) “Russia: Hacker Hell, Scourge of the RuNet—Global Voices,” Global Voices (blog), July 23, 2012, https://globalvoices.org/2012/07/23/russia-hacker-hell-scourge-of-the-runet/; Aleksandr Gorbachev On 7/9/15 at 6:43 AM, “Meet the Hacker Who Terrorized the Russian Blogosphere,” Newsweek, July 9, 2015, http://www.newsweek.com/2015/07/17/gospel-according-hell-351544.html; Sergey Sanovich, “Computational Propaganda in Russia: The Origins of Digital Misinformation,” n.d., 26.

(13.) Nakashima, “Russian Government Hackers Penetrated DNC, Stole Opposition Research on Trump,” Washington Post, June 14, 2016, sec. National Security, https://www.washingtonpost.com/world/national-security/russian-government-hackers-penetrated-dnc-stole-opposition-research-on-trump/2016/06/14/cf006cb4-316e-11e6-8ff7-7b6c1998b7a0_story.html.

(14.) “Can a BEAR Fit Down a Rabbit Hole? State Board of Election Analysis,” ThreatConnect | Enterprise Threat Intelligence Platform, September 2, 2016, https://www.threatconnect.com/blog/state-board-election-rabbit-hole/; “ThreatConnect Identifies FANCY BEAR World Anti-Doping Agency Breach,” ThreatConnect | Enterprise Threat Intelligence Platform, August 19, 2016, https:// (p.418) www.threatconnect.com/blog/fancy-bear-anti-doping-agency-phishing/; “Shiny Object? Guccifer 2.0 and the DNC Breach,” ThreatConnect | Enterprise Threat Intelligence Platform, June 29, 2016, https://www.threatconnect.com/blog/guccifer-2-0-dnc-breach/; “ThreatConnect Identifies Additional Infrastructure in DNC Breach,” ThreatConnect | Enterprise Threat Intelligence Platform, June 17, 2016, https://www.threatconnect.com/blog/tapping-into-democratic-national-committee/; “APT28: A Window into Russia’s Cyber Espionage Operations?,” FireEye, 28, accessed March 13, 2018, https://www.fireeye.com/blog/threat-research/2014/10/apt28-a-window-into-russias-cyber-espionage-operations.html.

(15.) “Joint Statement from the Department Of Homeland Security and Office of the Director of National Intelligence on Election Security” (Department of Homeland Security, October 7, 2016), https://www.dhs.gov/news/2016/10/07/joint-statement-department-homeland-security-and-office-director-national; “Joint Analysis Report: GRIZZLY STEPPE—Russian Malicious Cyber Activity” (Department of Homeland Security and Federal Bureau of Investigation, December 29, 2016), https://www.us-cert.gov/sites/default/files/publications/JAR_16-20296A_GRIZZLY%20STEPPE-2016-1229.pdf.

(16.) “Assessing Russian Activities and Intentions in Recent US Elections,” Intelligence Community Assessment, January 6, 2017, https://www.dni.gov/files/documents/ICA_2017_01.pdf.

(17.) Indictment, United States of America vs. Internet Research Agency LLC et al. 2/16/18. https://www.justice.gov/file/1035477/download.

(18.) Jen Weedon, William Nuland, and Alex Stamos, “Information Operations and Facebook” (Facebook, April 27, 2017), https://fbnewsroomus.files.wordpress.com/2017/04/facebook-and-information-operations-v1.pdf; Alex Stamos, “An Update On Information Operations On Facebook” (Facebook, September 6, 2017), https://newsroom.fb.com/news/2017/09/information-operations-update/.

(19.) “Update: Russian Interference in 2016 US Election, Bots, & Misinformation” (Twitter, September 28, 2017), https://blog.twitter.com/official/en_us/topics/company/2017/Update-Russian-Interference-in-2016--Election-Bots-and-Misinformation.html.

(20.) Young Mie Kim, Jordan Hsu, David Neiman, Colin Kou, Levi Bankston, Soo Yun Kim, Richard Heinrich, Robyn Baragwanath, and Garvesh Raskutt, “The Stealth Media? Groups and Targets Behind Divisive Issue Campaigns on Facebook.” Forthcoming Political Communication. Available at https://journalism.wisc.edu/wp-content/blogs.dir/41/files/2018/04/Kim.FB_.StealthMedia.re_.3.two-colmns.041718-1.pdf.

(21.) Mike Isaac and Daisuke Wakabayashi, “Russian Influence Reached 126 Million Through Facebook Alone,” New York Times, October 30, 2017, sec. Technology, https://www.nytimes.com/2017/10/30/technology/facebook-google-russia.html; Colin Lecher, “Here Are the Russia-Linked Facebook Ads Released by Congress,” The Verge, November 1, 2017, https://www.theverge.com/2017/11/1/16593346/ (p.419) house-russia-facebook-ads; Nicholas Fandos, Cecilia Kang, and Mike Isaac, “House Intelligence Committee Releases Incendiary Russian Social Media Ads,” New York Times, November 1, 2017, sec. Politics, https://www.nytimes.com/2017/11/01/us/politics/russia-technology-facebook.html.

(22.) Ben Collins, Kevin Poulsen, and Spencer Ackerman, “Exclusive: Russia Used Facebook Events to Organize Anti-Immigrant Rallies on U.S. Soil,” Daily Beast, September 12, 2017, sec. Tech, https://www.thedailybeast.com/exclusive-russia-used-facebook-events-to-organize-anti-immigrant-rallies-on-us-soil.

(23.) Georgia Wells, “Russia-Linked Facebook Pages Pushed Divisions After Election, Including on Charlottesville,” Wall Street Journal, October 3, 2017, sec. Tech, https://www.wsj.com/articles/russia-linked-facebook-pages-pushed-divisive-views-through-august-1507051387.

(24.) Scott Shane, “The Fake Americans Russia Created to Influence the Election,” New York Times, September 7, 2017, sec. Politics, https://www.nytimes.com/2017/09/07/us/politics/russia-facebook-twitter-election.html.

(25.) @DFRLab, “The Russians Who Exposed Russia’s Trolls.”

(26.) “Russia Media Group That Angered Kremlin Is Sold,” Reuters, June 16, 2017, https://www.reuters.com/article/us-russia-media-rbc-sale/russia-media-group-that-angered-kremlin-is-sold-idUSKBN1972KO.

(27.) “Расследование РБК: Как «фабрика Троллей» Поработала На Выборах в США,” Журнал РБК, accessed March 13, 2018, https://www.rbc.ru/magazine/2017/11/59e0c17d9a79470e05a9e6c1.

(28.) Whitney Phillips, Jessica Beyer, and Gabriella Coleman, “Trolling Scholars Debunk the Idea That the Alt-Right’s Shitposters Have Magic Powers,” Motherboard, March 22, 2017, https://motherboard.vice.com/en_us/article/z4k549/trolling-scholars-debunk-the-idea-that-the-alt-rights-trolls-have-magic-powers.

(29.) @DFRLab, “The Russians Who Exposed Russia’s Trolls.”

(30.) Denis Stukal et al., “Detecting Bots on Russian Political Twitter,” Big Data 5, no. 4 (December 1, 2017): 310–324, https://doi.org/10.1089/big.2017.0038; @DFRLab, “Human, Bot or Cyborg?,” DFRLab (blog), December 23, 2016, https://medium.com/@DFRLab/human-bot-or-cyborg-41273cdb1e17.

(31.) @DFRLab, “The Russians Who Exposed Russia’s Trolls.”

(32.) Jane Lytvynenko, “InfoWars Has Republished More Than 1,000 Articles From RT Without Permission,” BuzzFeed, accessed March 13, 2018, https://www.buzzfeed.com/janelytvynenko/infowars-is-running-rt-content.

(33.) “Sean Adl-Tabatabai on Being in the Eye of the ‘Fake News’ Storm | London Evening Standard,” accessed March 13, 2018, https://www.standard.co.uk/lifestyle/london-life/sean-adltabatabai-on-being-in-the-eye-of-the-fake-news-storm-a3468361.html.

(34.) Kate Starbird, “Created a Domain Network Graph for ‘White Helmets’ Tweets May–Aug 2017. These Are Websites Connected by Users (Who Tweet to Both).,” Twitter (blog), September 21, 2017, https://twitter.com/katestarbird/ (p.420) status/910988793125023745?lang=en; “Killing the Truth” (The Syria Campaign, December 20, 2017), http://thesyriacampaign.org/wp-content/uploads/2017/12/KillingtheTruth.pdf.

(37.) Adam Carter, “Guccifer 2.0: Game Over,” Guccifer 2.0, February 22, 2018, http://g-2.space.

(38.) нет, “Russia and WikiLeaks: The Case of the Gilded Guccifer,” Medium (blog), February 21, 2017, https://web.archive.org/web/20170221231421/ https://medium.com/@nyetnyetnyet/russia-and-wikileaks-the-case-of-the-gilded-guccifer-f2288521cdee.

(39.) нет, “This Fancy Bear’s House Is Made of Cards: Russian Fools or Russian Frame-Up?,” Medium (blog), March 8, 2017, https://web.archive.org/web/20170308190211/https://medium.com/@nyetnyetnyet/this-fancy-bears-house-is-made-of-cards-russian-fools-or-russian-frame-up-59a714243b91.

(40.) “Intelligence Committee Ranking Member Schiff Opening Statement During Hearing on Russian Active Measures,” U.S. House of Representatives Permanent Selection on Intelligence Democratic Office, March 20, 2017, https://democrats-intelligence.house.gov/news/documentsingle.aspx?DocumentID=220.

(41.) Cassandra Fairbanks, “Pretty Sure Guccifer 2 Is Not A ‘creature of Russia’ @RepAdamSchiff,” Twitter, March 20, 2017, 2, https://twitter.com/cassandrarules/status/843853011562708992?lang=en.

(42.) “Twitter DMs—Guccifer 2.0,” Guccifer 2.0, n.d., http://g-2.space/sr/dms.html.

(43.) WikiLeaks, “Direct Messages from U.S. Alleged Russian Spy @GUCCIFER_2 to Actress-Model @robbin_young (according to the Latter),” Twitter, April 8, 2017, https://twitter.com/wikileaks/status/850692885062139904.

(44.) Tyler Durden, “‘Guccifer 2.0’ Chat With Nude Model Sparks New Conspiracy Theories About Seth Rich Murder,” Zero Hedge, April 9, 2017, https://www.zerohedge.com/news/2017-04-09/guccifer-20-chat-nude-model-sparks-new-conspiracy-theories-about-seth-rich-murder.

(45.) Jim Hoft, “Report: Wikileaks Bombshell: Guccifer 2.0 Says ‘Seth’ Rich Was DNC Leaker,” Gateway Pundit, April 8, 2017, http://www.thegatewaypundit.com/2017/04/wikileaks-bombshell-guccifer-2-0-admits-seth-rich-dnc-leaker/.

(46.) Crooked Hillary, “Fox News Report on Messages between Guccifer 2.0 and Robin Young Regarding Seth Rich” (YouTube, 2017), https://www.youtube.com/watch?v=s5h28kIcEAk.

(47.) Cassandra Fairbanks, “My Strange Interactions With Guccifer 2.0,” Big League Politics, April 23, 2017, https://bigleaguepolitics.com/strange-interactions-guccifer-2-0/.

(48.) Cassandra Fairbanks, “Hacker: Guccifer 2.0 Planted Fake Russian Fingerprints On His Account,” Big League Politics, April 25, 2017, https://bigleaguepolitics.com/hacker-guccifer-2-0-planted-fake-russian-fingerprints-account/.

(49.) Jo Becker, Matt Apuzzo, and Adam Goldman, “Trump Team Met With Lawyer Linked to Kremlin During Campaign,” New York Times, July 8, 2017, sec. Politics, https://www.nytimes.com/2017/07/08/us/politics/trump-russia-kushner-manafort.html.

(50.) “Guccifer 2.0 NGP/VAN Metadata Analysis,” The Forensicator (blog), July 9, 2017, https://theforensicator.wordpress.com/guccifer-2-ngp-van-metadata-analysis/.

(51.) Elizabeth Vos, “New Research Shows Guccifer 2.0 Files Were Copied Locally, Not Hacked,” Disobedient Media, July 9, 2017, https://disobedientmedia.com/2017/07/new-research-shows-guccifer-2-0-files-were-copied-locally-not-hacked/.

(52.) “William Craddick | Zero Hedge,” Zero Hedge, accessed April 5, 2018, https://www.zerohedge.com/users/william-craddick.

(53.) Zeiger, “New Analysis Disproves Russian Hacking Hoax Narrative,” Daily Stormer, July 22, 2017, https://web.archive.org/web/20170722182420/https://www.dailystormer.com/new-analysis-disproves-russian-hacking-hoax-narrative/.

(54.) “New Analysis Suggests Guccifer 2.0 Files Copied Locally, Not Hacked by Russia,” RT International, July 12, 2017, https://www.rt.com/usa/396056-guccifer-forensicator-hack-dnc/.

(55.) veganmark, “Implications of Recent Analyses by Adam Carter and Forensicator,” Reddit, July 10, 2017, https://www.reddit.com/r/WayOfTheBern/comments/6mgjuy/implications_of_recent_analyses_by_adam_carter/.”

(56.) “Intel Vets Challenge ‘Russia Hack’ Evidence,” Consortiumnews (blog), July 24, 2017, https://consortiumnews.com/2017/07/24/intel-vets-challenge-russia-hack-evidence/.

(57.) Nicholas D. Kristof, “Save Our Spooks,” New York Times, May 30, 2003, sec. Opinion, https://www.nytimes.com/2003/05/30/opinion/save-our-spooks.html.

(58.) Patrick Lawrence, “A New Report Raises Big Questions About Last Year’s DNC Hack,” The Nation, August 9, 2017, https://www.thenation.com/article/a-new-report-raises-big-questions-about-last-years-dnc-hack/.

(59.) Leonid Bershidsky, “Why Some U.S. Ex-Spies Don’t Buy the Russia Story,” Bloomberg.com, August 10, 2017, https://www.bloomberg.com/view/articles/2017-08-10/why-some-u-s-ex-spies-don-t-buy-the-russia-story.

(60.) Brian Feldman, “The Nation Article About the DNC Hack Is Too Incoherent to Even Debunk,” Select All, August 10, 2017, http://nymag.com/selectall/2017/08/the-nation-article-about-the-dnc-hack-is-incoherent.html; Nancy LeTourneau, “Conspiracy Theories on the Left,” Washington Monthly, August 11, 2017, https://washingtonmonthly.com/2017/08/11/conspiracy-theories-on-the-left/.

(61.) Danielle Ryan, “What If the DNC Russian ‘hack’ Was Really a Leak after All? A New Report Raises Questions Media and Democrats Would Rather Ignore,” August 15, 2017, https://www.salon.com/2017/08/15/what-if-the-dnc-russianhack-was-really-a-leak-after-all-a-new-report-raises-questions-media-and-democrats-would-rather-ignore.

(62.) Ian Mason, “Left-Wing Magazine The Nation Report Puts ‘Russian Hack’ DNC Narrative in Freefall,” Breitbart, August 10, 2017, http://www.breitbart.com/ (p.422) big-government/2017/08/10/left-wing-magazine-the-nation-report-puts-russian-hack-dnc-narrative-in-freefall/.

(63.) “Former NSA Official: Dems’ Russia Hacking Story Likely Bogus,” Fox News Insider, August 14, 2017, http://insider.foxnews.com/2017/08/14/russian-hacking-2016-election-nsa-member-says-trump-story-false.

(64.) Erik Wemple, “The Nation Is Reviewing a Story Casting Doubt on Russian Hack of DNC,” Washington Post, August 15, 2017, sec. Opinion, https://www.washingtonpost.com/blogs/erik-wemple/wp/2017/08/15/the-nation-is-reviewing-a-story-casting-doubt-on-russian-hack-of-dnc/.

(65.) Various Contributors, “A Leak or a Hack? A Forum on the VIPS Memo: Independent Review,” The Nation, September 1, 2017, https://www.thenation.com/article/a-leak-or-a-hack-a-forum-on-the-vips-memo/#independent-review.

(66.) Nathaniel Freitas, “Independent Review of Reporting and Analysis on the 2016 Compromise of the Dnc Computer Network,” The Nation, September 1, 2017, https://www.thenation.com/article/a-leak-or-a-hack-a-forum-on-the-vips-memo/#independent-review.

(67.) THOMAS DRAKE, SCOTT RITTER, LISA LING, CIAN WESTMORELAND, PHILIP M. GIRALDI, AND JESSELYN RADACK, THE NATION, September 1, 2017, WHEN FACTS ARE NOT FACTS,https://www.thenation.com/article/a-leak-or-a-hack-a-forum-on-the-vips-memo/.

(68.) William Binney, Skip Folden, Ed Loomis, Ray Mcgovern, and Kirk Wiebe, “Why This Is Important,” The Nation, September 1, 2017, When Facts are not Facts, https://www.thenation.com/article/a-leak-or-a-hack-a-forum-on-the-vips-memo/.

(69.) Oren Dorell, “Breitbart, Other ‘alt-right’ Websites are the Darlings of Russian Propaganda Effort,” USA Today, August 24, 2017, https://www.usatoday.com/story/news/world/2017/08/24/breitbart-other-alt-right-websites-darlings-russian-propaganda-effort/598258001/.

(70.) Bethania Palma, “Was the Manchester Terror Attack a ‘False Flag’?,” Snopes.com, May 24, 2017, https://www.snopes.com/fact-check/manchester-attack-false-flag/.

(71.) “All Statements Involving YourNewsWire.Com | PolitiFact,” PolitiFact, accessed April 6, 2018, http://www.politifact.com/personalities/yournewswirecom/statements/?list=speaker.

(72.) “L.A. Alt-Media Agitator (Not Breitbart) Clashes With Google, Snopes,” Hollywood Reporter, accessed March 13, 2018, https://www.hollywoodreporter.com/features/hollywoods-hidden-alt-media-firebrands-1041157.

(73.) Richard Godwin, “Sean Adl-Tabatabai on Being in the Eye of the ‘Fake News’ Storm,” London Evening Standard, February 16, 2017, https://www.standard.co.uk/lifestyle/london-life/sean-adltabatabai-on-being-in-the-eye-of-the-fake-news-storm-a3468361.html; Josh Boswell, “Mother Churns out Stories for Master of Fake News,” Sunday Times, January 29, 2017, sec. News, https://www.thetimes.co.uk/article/mother-churns-out-stories-for-master-of-fake-news-fcmzc05sx.

(74.) Tracy Alloway and Luke Kawa, “Unmasking the Men Behind Zero Hedge, Wall Street’s Renegade Blog,” Bloomberg, April 29, 2016.

(75.) “Re: No Shit,” email, WikiLeaks, May 5, 2016, https://wikileaks.org/dnc-emails/emailid/11508.

(76.) Brent Johnson, “The 12 Questions and Answers of Donald Trump’s Reddit AMA,” NJ.com, July 27, 2016, http://www.nj.com/politics/index.ssf/2016/07/the_9_questions_and_answers_of_donald_trumps_reddit_ama.html.

(77.) Associated Press, “Republican-Authored Voting Laws in Wisconsin and Kansas Overturned,” The Guardian, July 29, 2016, http://www.theguardian.com/us-news/2016/jul/29/wisconsin-kansas-republican-voting-laws-unconstitutional.

(78.) Charlie Nash, “Roger Stone on The Milo Show: How Trump Can Fight Voter Fraud,” Breitbart, July 29, 2016, http://www.breitbart.com/milo/2016/07/29/roger-stone-milo-show-trump-can-fight-voter-fraud/.

(79.) Katherine Krueger, “Trump: ‘I’m Afraid’ The General Election’s ‘Gonna Be Rigged,’ ” TPM, August 1, 2016, https://web.archive.org/web/20160803151821/http:/talkingpointsmemo.com:80/livewire/trump-general-election-rigged-potential-challenge, Internet Archive record.

(80.) Allan Smith, “DONALD TRUMP: ‘I’m Afraid the Election Is Going to Be Rigged,’ ” Business Insider, August 1, 2016, http://www.businessinsider.com/donald-trump-rigged-election-2016-8.

(81.) “Donald Trump: The Election in November Will Be Rigged,” video, Fox News, August 1, 2016, http://video.foxnews.com/v/5067615849001/.

(82.) Ian Schwartz, “Trump: People Without Identification ‘Are Going To Vote 10 Times,’ ” video, RealClearPolitics, August 2, 2016, https://www.realclearpolitics.com/video/2016/08/02/trump_people_are_going_to_vote_10_times.html.

(83.) A.J. Katz, “August 2016 Ratings: Fox News Is Most-Watched Cable Network for a Second Straight Month,” TVNewser, August 30, 2016, http://adweek.it/2bGKntn.

(84.) “Why Is Trump Saying the Election Is Rigged?,” Rush Limbaugh Show (blog), August 2, 2016, https://www.rushlimbaugh.com/daily/2016/08/02/why_is_trump_saying_the_election_is_rigged_/.

(86.) “Pelosi, Reid Call for All Leaders to Join in Affirming Fairness of Our Democratic Process,” Democratic Leader Nancy Pelosi, October 17, 2016, https://www.democraticleader.gov/newsroom/pelosi-reid-call-for-all-leaders-to-join-in-affirming-fairness-of-our-democratic-process/.

(87.) Zachary Roth, “Analysis: Trump Not the First to Claim Voter Fraud Will Rig Elections—NBC News,” NBC News, August 3, 2016, https://www.nbcnews.com/politics/2016-election/trump-not-first-claim-voter-fraud-will-rig-elections-n622421.

(88.) Katherine Q. Seelye, “McCain’s Warning About Voter Fraud Stokes a Fiery Campaign Even Further—The New York Times,” New York Times, October 26, 2008, https://www.nytimes.com/2008/10/27/us/politics/27vote.html?_r=0.

(89.) Tom Jensen, “ACORN,” Public Policy Polling (blog), November 19, 2009, https://publicpolicypolling.blogspot.com/2009/11/acorn.html.

(90.) “2016 Election Results: Michigan Exit Polls,” CNN, November 9, 2016, http://www.cnn.com/election/2016/results/exit-polls/michigan/president.

(91.) Joshua Green and Sasha Issenberg, “Inside the Trump Bunker, With 12 Days to Go,” Bloomberg, October 27, 2016, https://www.bloomberg.com/news/articles/2016-10-27/inside-the-trump-bunker-with-12-days-to-go.

(92.) Daniel Kreiss and Shannon C. Mcgregor, “Technology Firms Shape Political Communication: The Work of Microsoft, Facebook, Twitter, and Google With Campaigns During the 2016 U.S. Presidential Cycle,” Political Communication, October 26, 2017, 1–23, https://doi.org/10.1080/10584609.2017.1364814.

(94.) “Alex ‘Never Giving Up’ Jones Gives Up,” accessed March 13, 2018, https://www.thedailybeast.com/alex-never-giving-up-jones-gives-up.

(95.) Alex Riggins, Nathan Brown, and Julie Wootton, “UPDATED: Story of Syrian Refugees Raping Idaho Girl Is Wrong, Prosecutor Says | Idaho Statesman,” Idaho Statesman, June 20, 2016, http://www.idahostatesman.com/news/state/idaho/article84829787.html; Rocky Barker, “Breitbart’s Chobani Coverage Is a Reminder of Idaho’s Experience with the Alt-Right,” Idaho Statesman, November 16, 2016, http://www.idahostatesman.com/news/local/news-columns-blogs/letters-from-the-west/article115283313.html.

(96.) Scott Shane, “Purged Facebook Page Tied to the Kremlin Spread Anti-Immigrant Bile,” New York Times, September 12, 2017, sec. Politics, https://www.nytimes.com/2017/09/12/us/politics/russia-facebook-election.html.

(97.) Bence Kollanyi, Philip N. Howard, and Samuel C. Woolley, “Bots and Automation over Twitter during the U.S. Election,” Computational Propaganda Research Project (Oxford Internet Institute, November 17, 2016), 1–4, http://comprop.oii.ox.ac.uk/wp-content/uploads/sites/89/2016/11/Data-Memo-US-Election.pdf.

(98.) Onur Varol et al., “Online Human-Bot Interactions: Detection, Estimation, and Characterization,” ArXiv Preprint ArXiv:1703.03107, 2017, https://arxiv.org/abs/1703.03107; Clayton Allen Davis et al., “BotOrNot: A System to Evaluate Social Bots,” in Proceedings of the 25th International Conference Companion on World Wide Web, WWW ‘16 Companion (Republic and Canton of Geneva, Switzerland: International World Wide Web Conferences Steering Committee, 2016), 273–274, https://doi.org/10.1145/2872518.2889302; Alessandro Bessi and Emilio Ferrara, “Social Bots Distort the 2016 U.S. Presidential Election Online Discussion,” First Monday 21, no. 11 (November 3, 2016), http://firstmonday.org/ojs/index.php/fm/article/view/7090.

(99.) Stukal et al., “Detecting Bots on Russian Political Twitter.”

(100.) Stefano Cresci et al., “The Paradigm-Shift of Social Spambots: Evidence, Theories, and Tools for the Arms Race,” in Proceedings of the 26th International Conference on World Wide Web Companion (International World Wide Web Conferences Steering Committee, 2017), 963–972, http://dl.acm.org/citation.cfm?id=3055135.

(101.) In full disclosure, the email reports that experts were consulted, including at the Berkman Center at Harvard, where all three of us were then, in the preparation of this memo. None of us remember whether we were consulted, but Yochai later wrote a detailed article explaining why WikiLeaks had to be treated no differently from the New York Times or the Guardian, which had also published the war logs and cables, and then participated in the defense of Chelsea Manning’s prosecution for aiding the enemy.

(102.) “Under Intense Pressure to Silence Wikileaks, Secretary of State Hillary Clinton Proposed Drone Strike on Julian Assange,” True Pundit (blog), October 2, 2016, https://truepundit.com/under-intense-pressure-to-silence-wikileaks-secretary-of-state-hillary-clinton-proposed-drone-strike-on-julian-assange/.

(103.) “Kremlin-Linked Network Amplifies ‘alt-Right’ Media.”