Right-Wing Sting Group Project Veritas Is Breaking Facebook’s “Authentic Behavior” Rule. Now What?

A member of Project Veritas gave testimony in a federal court case indicating that the right-wing group, known for its undercover videos, violates Facebook policies designed to counter systematic deception by Russian troll farms and other groups. The deposition raises questions over whether Facebook will deter American operatives who use the platform to strategically deceive and damage political opponents as vigorously as it has Iranian and Russian propagandists. But is the company capable of doing so without just creating more problems?

Close observers of Veritas and Facebook, including one at a research lab that works with the social network, said the testimony shows the group is clearly violating policies against what Facebook refers to as “coordinated inauthentic behavior.” The company formally defined such behavior in a December 2018 video featuring its cybersecurity policy chief Nathaniel Gleicher, who said it “is when groups of pages or people work together to mislead others about who they are or what they’re doing.” The designation, Gleicher added, is applied by Facebook to a group not “because of the content they’re sharing” but rather only “because of their deceptive behavior.” That is, using Facebook to dupe people is all it takes to fit the company’s institutional definition of coordinated inauthentic behavior.

In practice, “coordinated inauthentic behavior” has become a sort of catchall label for untoward meddling on Facebook, snagging everyone from Burmese military officers to Russian meme spammers. But curbing such activity has become a very public crusade for Facebook in the wake of its prominent role as a platform for the spread of disinformation, propaganda, and outright hoaxes during the 2016 presidential campaign. This past January, Gleicher announced the removal of coordinated inauthentic behavior from Iran, which spread when operatives “coordinated with one another and used fake accounts to misrepresent themselves,” thus triggering a Facebook ban. Similarly, in a 2017 update on Facebook’s internal investigation into Russian online propaganda efforts, the company’s then-head of security Alex Stamos assured the world’s democracies the company was providing “technology improvements for detecting fake accounts,” including “changes to help us more efficiently detect and stop inauthentic accounts at the time they are being created.”

Throughout all of this, coordinated inauthentic behavior has remained more or less synonymous with “foreign actors” and “nation-states,” the cloak-and-dagger stuff of an increasingly militarized internet filled with enemies of the Western Democracy who seek to subvert it from abroad.

Project Veritas, a hybrid of an opposition research shop and a ranting YouTube channel, has taken pride in its ability to deceive since its creation in 2010. With conservative backers like Peter Thiel, the Koch brothers, and the Trump Foundation, the group and its founder James O’Keefe have worked relentlessly to target and malign individuals at institutions they deem leftist, whether it’s Planned Parenthood (reportedly targeted by O’Keefe posing as a young teen’s 23-year-old boyfriend), George Soros (the progressive philanthropist whose professional circle Veritas tried and spectacularly failed to infiltrate), or the Washington Post (whose reporter was offered a fake story on Alabama Senate candidate Roy Moore). O’Keefe has long attempted to position himself in the context of dogged, daring, traditional journalism, describing Veritas’s efforts as “investigative” reporting executed by “undercover journalists.” But his efforts are often executed by what the New Yorker has called “amateurish spies” — their efforts against the Post and Soros resembled a Three Stooges bit — and packaged with mendacious editing, duplicitous production, and outright lying, making Veritas’s audience as much a victim of its productions as the subjects. Debates over who or what is to be considered “real journalism” are almost always counterproductive and contrived, but Veritas stands out for the shamelessness with which it pursues nakedly partisan ends.

There is, of course, a proud tradition of undercover journalism executed unequivocally in the name of informing the public. Writers like Barbara Ehrenreich and Shane Bauer have taken jobs they were not otherwise interested in in order to reveal injustices in society’s margins, and some of the most damning details of the Cambridge Analytica scandal were exposed by a reporter with the UK’s Channel 4 posing as a foreign politician interested in the company’s services. This reporting involved lying, sure — or at least the withholding of true intent, and a willingness to let others deceive themselves — but only as a means to a truthful end. The distinction between these reporters and Veritas operatives may be that the end the latter group seeks, the final media product, is typically just another act of partisan misdirection that doesn’t withstand further scrutiny.

Neither Project Veritas nor Facebook commented for this story.

“Legend Building” by Project Veritas

Project Veritas has systematically deceived not just targets on the left and viewers on the right but Facebook users as well (their official page has over 200,000 followers) at a time when the company is publicly dedicated to fighting this sort of systemic duplicity. That’s a wrinkle that raises questions about Facebook’s commitment to rooting out coordinated inauthentic behavior closer to home — Thiel sits on the company’s board — not to mention Project Veritas’s presence on social media.

“We thus have the admission of intent by the organization and evidence of action by multiple of its agents.”

In 2017, O’Keefe sued the Suffolk County district attorney over a Massachusetts law barring the covert recording of government officials. This past December, a federal judge overturned the rule. But in the course of the lawsuit, Joe Halderman, a member of the Project Veritas inner circle who was previously convicted of trying to extort late night television host David Letterman in 2010, sat for a deposition. In it, Halderman was compelled to submit to a sworn interrogation of Veritas methods. Just how does one go about duping savvy politicos and the politico-adjacent in the 21st century?

During his deposition, Halderman, Project Veritas’s self-described “executive producer,” stated under oath that the organization falsifies Facebook accounts as part of its overall strategy of deceiving the targets of its investigations. Halderman, characterizing himself as “integrally involved in [Project Veritas’s] investigations and have been since I started four years ago,” describes the work that went into setting up Robert Creamer, a Democratic operative recorded by Veritas in a 2016. That video attempted to portray Creamer as complicit in a Hillary Clinton-led campaign to violently disrupt Donald Trump’s campaign events with counterprotests and engaging in counter-Trump voter fraud — both regular, unfounded talking points repeated by Trump on the campaign trail. Last year, the Wisconsin Department of Justice concluded an investigation into the videos, determining that they “reveal no evidence of election fraud,” the Associated Press reported.

But before Veritas could get Creamer on camera, they needed to make contact via a fake persona, Halderman explained. Per the transcript (emphasis added):

Q. And you spoke earlier about PVA [Veritas] creating this donor, Charles. When you say create the donor, what did PVA do to create the donor?

A. So, I thought of a name. I talked to the undercover journalist who was the person who met with Foval. We between us sort of created this story of this person. I got some business cards made. I got an e-mail. I set up an e-mail account. What else did I do? I think that’s about all I did.

Again, in this particular case, we didn’t feel like they were going to get seriously vetted. In some investigations we do legend building because we believe or our concern is that we’re going to be vetted reasonably, you know, by open source information.

So, we’ll create a Facebook page, a LinkedIn page. We’ve even gone so far in the past of creating LLCs, offshore bank accounts. We do a lot of things because undercover journalism is a tricky, complicated business.

According to Lauren Windsor, a political organizer and partner at Democracy Partners (alongside Robert Creamer) who began documenting Veritas’s team of “undercover” operatives and their various aliases after her own organization was infiltrated, this sort of use of phony social accounts is the group’s standard operating procedure. “In conducting extensive outreach to victims and extensive research of social media networks to build the vetting resource website Project Veritas Exposed,” explained Windsor, “I documented several instances of PV violating Facebook’s terms by creating fake profiles. We thus have the admission of intent by the organization and evidence of action by multiple of its agents.”

Windsor’s work includes cataloging Project Veritas’s network of fake Facebook accounts; Windsor provided screenshots to The Intercept. In one example, Veritas operative Marisa Jorge’s likeness is used for the Facebook profile of “Ava-Marie Joyce.” The bio of the Ava-Marie persona bizarrely describes herself as “Carrie Tallinn, a self-employed professional women’s right activist.”

According to a 2018 lawsuit reported by The Intercept last year, Jorge previously misrepresented herself as a University of Michigan student in order to gain improper access to teachers union documents. The friends list for “Ava-Marie Joyce” lists another profile fabricated by Project Veritas, “Ava Marie Allen.”

Another screenshot shows a Facebook profile for “Tyler Marshall,” which O’Keefe himself disclosed as a fabricated identity in his 2018 book “American Pravda.” In a section of that book (subtitle: “My Fight For Truth in the Era of Fake News”) detailing Veritas’s attempts to infiltrate protestors planning action around Trump’s presidential inauguration, O’Keefe wrote that his operatives all “of course, establish a social media presence under their assumed names—‘Tyler Marshall,’ say, or ‘Adam Stevens.’ The presence includes Twitter, Facebook, and email at the least.”

Exploitation of Facebook by a Group Linked To Military Intelligence

Emerson Brooking, a resident fellow at the Atlantic Council’s Digital Forensics Research Lab, said Veritas’s homegrown social media deception is a clear violation of Facebook’s policy. After reading the deposition, Brooking told The Intercept, “Mr. Halderman describes the creation of fake Facebook personas for the purpose of deception” and “implies that this is a regular and systematic practice. Under any reasonable definition, Project Veritas is engaged in coordinated inauthentic behavior and abuse of the Facebook platform.”

“Under any reasonable definition, Project Veritas is engaged in coordinated inauthentic behavior and abuse of the Facebook platform.”

Brooking’s group, a frequently cited authority on online electoral interference and other digital propaganda campaigns, entered into an official partnership with Facebook last year. A Facebook director wrote at the time that experts at the lab “will work closely with our security, policy and product teams to get Facebook real-time insights and updates on emerging threats and disinformation campaigns from around the world.”

If it sounds like a stretch to compare Project Veritas to a Russian troll farm, consider the group’s links to the U.S. defense establishment. As The Intercept reported in May, Veritas members underwent “intelligence and elicitation techniques from a retired military intelligence operative named Euripides Rubio Jr.,” personally arranged by the infamous American mercenary and Trump adviser Erik Prince. What we have here, then, is a 2016 military intelligence-linked, organized effort to undermine the Democratic Party and boost the Trump presidential campaign using falsified social media profiles. If that doesn’t sound familiar, it certainly should.

The problem with Facebook and its peers has never been identifying abuses and misuses, whether truly dangerous or merely toxic; Facebook, Twitter, and Google alone represent perhaps history’s greatest living catalog of antisocial behavior, a frenzy of rule violation on a mass scale. Whether these companies deem comprehensive content moderation simply too expensive or not worth the public relations mess, the fact is that the public rarely sees movement on these issues in the absence of congressional scolding or media uproar.

The real issue is uneven, arbitrary enforcement of “the rules.” Max Read, writing in New York magazine on another social network’s enforcement blunders, argued that “the problem for YouTube is that for rules to be taken seriously by the people they govern, they need to be applied consistently and clearly.” YouTube is about as terrible at this exercise as Facebook is, and there’s a good chance that if Facebook treated malicious right-wing American exploitation of its network the same way it treats malicious foreign exploitation of its network, it would probably botch the whole thing and end up burning people who actually do use phony Facebook profiles for work toward the public good.

That a company like Facebook is even in a position to create “rules” like the coordinated inauthentic behavior policy that apply to a large chunk of the Earth’s population is itself a serious problem, one made considerably worse by completely erratic enforcement. It’s bad enough having a couple guys in California take up the banner of defending “Democracy” around the world through the exclusive control of one of the most powerful information organs in human history; if nothing else, we should hope their decisions are predictable and consistent.

Correction: June 11th, 2019, 11:19 a.m.

This article has been updated to name Lauren Windsor’s employer, Democracy Partners, where she is a partner.