What Were the Twitter Files?

What Were the Twitter Files?

While it’s clear Matt Taibbi made some serious factual errors, criticisms of the disclosures have ignored the role of the FBI and other parts of the national security state.


What exactly were “the Twitter Files” about? By now, it’s settled into a near-consensus everywhere but on the right that the disclosures were of little consequence. This view only hardened after MSNBC anchor Mehdi Hasan challenged Matt Taibbi, the journalist most associated with the so-called Files, on several major reporting errors two weeks ago.

Taibbi had mixed up the timeline of the creation of the Election Integrity Partnership (EIP), the public-private project set up to monitor social media misinformation during the election, and had vastly overstated the number of tweets it had flagged for removal, from 2,890 tweets to 22 million—“Taibbi’s most consequential claim,” according to one commentator.

Most seriously, Hasan showed that in charging that the EIP “was partnered with state entities,” Taibbi had erroneously identified the Center for Internet Security (CIS), a private nonprofit—as the CISA, the Department of Homeland Security’s (DHS) Cybersecurity and Infrastructure Security Agency. Adding an “A” to the end of “CIS” allowed Taibbi to make what Hasan called the “false claim” that the EIP was partnered with the government.

As the MSNBC chyron asked whether there was “proof of [a] censorship regime,” viewers were clearly left to think there wasn’t. Since then, Hasan has said that Taibbi’s false identification of the CISA “was key to his thesis.” In a widely circulated post, Techdirt charged that it had been “a key linchpin in the argument that the government was sending tweets for Twitter to remove,” concluding that “there remains no there there.” Various pundits have since repeated these points, charging the entire “scary censorship narrative” has fallen apart.

For many, this was a gratifying end to the saga: A media figure that many viewed as having gone over to the Dark Side of the conservative culture wars was taken down on live television. But reducing the matter to an episode of media comeuppance does a profound public disservice when, despite Taibbi’s errors, the convergence of social media censorship and the national security establishment is both very real and deeply worrying.

And Taibbi’s critics’ overstatements are themselves deeply misleading. Take the issue of the CIS and the CISA. While Taibbi mixed up the two in his tweet, the fact that the CISA works with the EIP isn’t remotely a “false claim”: The EIP itself openly says its “partnership with CISA began under the Trump administration.”

Meanwhile, though the CIS certainly isn’t a government entity, it’s also received a little under $250 million of US government funding since 2008—the vast majority of it from the DHS. By the CIS’s own admission, both its divisions on election security and broader cybersecurity, which work together, are funded by grants from the very same CISA in question. It openly calls its cybersecurity division a “government entity” and boasts members from “all 50 states, 49 state capitals, as well as hundreds of local governments, tribal governments, and U.S. territories.” The CIS’s president and CEO is a former US Air Force and Department of Energy official who sat on the Cyber Security Commission under Barack Obama, while its other executives and board members hail from government entities like the National Security Agency (NSA), the Pentagon, and of course, the DHS. It even has a handy infographic showing the close relationship between it and the CISA.

There’s no question that the CIS is a private nonprofit. But it’s one so intertwined with the DHS, and so deeply connected to the similarly named CISA, that the lines between private and government are awfully blurry.

But in many ways, the CIS/CISA issue is beside the point, because this was by no means the most important revelation regarding government involvement in tech censorship. The three letters conspicuously missing from criticisms of the Twitter files are “F,” “B,” and “I,” with the disclosures having laid bare the alarming extent of the role that the Federal Bureau of Investigation now plays in the company’s “content moderation” policies.

Among the disclosures were the fact that the FBI was having monthly and even weekly meetings with Twitter executives to coordinate anti-misinformation efforts; that it was doing so in conjunction with entities like the DHS, the Department of Justice (DOJ), and the office of the Director of National Intelligence; that the Central Intelligence Agency and even, at one point, the NSA, were involved in such meetings; that in Twitter’s interactions with the FBI, former Head of Trust and Safety Yoel Roth considered the Bureau a proxy for the entire “intelligence community” as a whole; and that the FBI and Twitter have become closely enmeshed, through the voluminous hiring of former Bureau personnel, offers of temporary security clearances and classified information sharing, and the creation of special portals for it and other government agencies to flag content.

What did they do with this influence? Among other things, the FBI and other government entities regularly sent over lists and Excel spreadsheets of hundreds of problematic accounts and tweets, a deluge of censorship requests that Twitter employees were overwhelmed by; repeatedly pressured Twitter to find evidence of large-scale foreign disinformation that the company officials said didn’t exist; were involved in what one employee called “a sustained (If uncoordinated) effort by the I[ntelligence]C[ommunity] to push us to share more info & change our API policies,” an effort that involved congressional staffers; and that US officials simultaneously had Twitter exempt from censorship their own propaganda accounts. While critics charge that there’s no evidence of government censorship in the “files,” there are in fact numerous instances of employees acting on government requests, with Twitter’s former public policy chief acknowledging that it identifies misinformation via “the FBI/DHS.”

Twitter employees themselves didn’t share the blasé attitude of critics of the reporting. They were perturbed that the FBI was “just doing keyword searches for [terms of service] violations” when it flagged accounts. Even the company’s former deputy general counsel Jim Baker—himself a former FBI official—found it “odd that they are searching for violations of our policies.” Roth was uncomfortable with the implications of “state-controlled media” the Bureau’s increasingly aggressive demands brought up.

But you don’t need Twitter bigwigs to explain why this is all troubling. Take the FBI, which just in recent years has turned its spying power on Black activists protesting police brutality in ways that the ACLU has said are “built on anti-Black racial stereotypes,” investigated and infiltrated the Standing Rock protest movement, and carried out a nationwide sweep of Muslim households on the eve of the 2016 election. There is alarming evidence of far-right sympathies within the Bureau and its collaboration with far-right extremists for the purpose of targeting anti-fascists. The FBI most recently made headlines for its role in the prosecution of “Cop City” protesters in Atlanta and its surveillance of connected activist groups.

There’s no shortage of scandals in the bloated DHS either, from its surveillance of journalists, activists, and anti-Trump protesters, to the overlap between it and the membership of the far-right Oath Keepers, to the tune of hundreds of its current and former employees. This is all against the backdrop of a wider policy strategy involving the DHS and the DOJ that views left-leaning activist movements like environmentalism, animal rights, and anti-capitalism as potential domestic terrorism.

This bias is evident in the Twitter files. The FBI flagged tweets supporting the George Floyd protests as potentially “driven by foreign-controlled bots.” It likewise incorrectly flagged leftist Catalonian accounts as being of Russian origin, while Taibbi reported that a list of hundreds of supposedly Iranian government-linked accounts included a former newspaper reporter and the left-wing outlet Truthout. Intelligence reports flagged thousands of accounts “propagating anti-Bolsonaro/pro-Lula hashtags,” referring to the far-right former president of Brazil and his leftist challenger, respectively.

That Twitter often resisted the FBI isn’t especially encouraging either, since even it admitted it was fighting a losing battle. Commenting on Twitter’s resistance to one particular censorship request, one employee wrote that “our window on that is closing, given that government partners are becoming more aggressive.” When another employee “found no links to Russia” on a set of flagged accounts, they offered to “try to find a stronger connection” anyway.

Hasan played down these concerns by pointing out that Twitter officially acted on 40 percent of government requests, instead of 100 percent of them. This will be cold comfort to the many left-leaning users who have been on the other side of these requests. That includes the hundreds of Palestinian activists, journalists, and other users who have complained for years about being censored for alleged terms of service and other violations on the platform, which has become integral to Palestinian organizing. An internal report last year from Facebook, another platform that partners with the FBI to fight misinformation, determined that its policies “had an adverse human rights impact” on Palestinians’ democratic rights partly through “greater over-enforcement” against Arabic posts.

Little thought is being given to how such a censorship regime could be misused in the future. Indeed, national security agencies don’t tend to give up power and shrink but to accumulate it and grow. And the Twitter Files show ample evidence of the government already pushing for exactly that.

Since taking over Twitter, Musk has slashed the platform’s content moderation team, claiming that “the censorship bureau was let go.” But there is little reason to sleep soundly. Musk, whose businesses are highly reliant on government contracts, particularly from the Pentagon, has stopped Twitter’s practice of publishing regular transparency reports about government requests, raising the question of whether Musk has not so much ended the company’s cooperation with the national security state as simply hidden it from public view. Musk’s free speech crusade has not extended to restoring the account of the group responsible for the 2020 “BlueLeaks” disclosures about police misconduct, and he has willingly gone along with the Modi government’s censorship demands against critics in India, in line with his highly questionable definition of free speech as merely meaning censorship in accordance with whatever the law demands. Musk’s words of assurance aren’t credible either: in the same BBC interview in which he claimed to be resisting government censorship, he labeled as “total bullshit” the existence of a family emerald mine that he once openly boasted about.

But even if one holds the utmost faith in Musk’s free speech commitments, the Tesla billionaire won’t own Twitter forever. And the kind of government role in social media censorship laid bare in the Twitter files extends to a variety of other “industry partners” like Facebook, which has shown an equally alarming propensity for censoring content that deviates from official US policy.

Will progressives be comfortable when the speech that’s targeted for censorship doesn’t concern elections and vaccines but instead that which “undermines the public’s trust in the courts” and the financial system, or “provokes violence against key infrastructure”—all of which the CISA has plans to target? And given the cyclical nature of politics, do they trust Donald Trump or whichever other Republican inevitably ends up in the White House someday to use these powers responsibly, let alone to share their views of what constitutes online misinformation? Progressives should remember that objections to War on Terror excesses were largely about how the sprawling, invasive national security state created after 9/11 could easily be hijacked by an unscrupulous, authoritarian leader—a prophecy we got some taste of in Trump’s response to the protests of 2020.

Government control of what’s said on social media and the rest of the Internet is a hallmark of authoritarian systems that exist in countries like Russia and China. There’s good reason to worry that the enmeshment of social media platforms with the national security state outlined in the Twitter Files, if left unchecked, could lay the foundation for exactly that. Progressives will protest in outrage if and when this regime is weaponized by right-wing forces. But by then, it may prove too late.

Dear reader,

I hope you enjoyed the article you just read. It’s just one of the many deeply reported and boundary-pushing stories we publish every day at The Nation. In a time of continued erosion of our fundamental rights and urgent global struggles for peace, independent journalism is now more vital than ever.

As a Nation reader, you are likely an engaged progressive who is passionate about bold ideas. I know I can count on you to help sustain our mission-driven journalism.

This month, we’re kicking off an ambitious Summer Fundraising Campaign with the goal of raising $15,000. With your support, we can continue to produce the hard-hitting journalism you rely on to cut through the noise of conservative, corporate media. Please, donate today.

A better world is out there—and we need your support to reach it.


Katrina vanden Heuvel
Editorial Director and Publisher, The Nation

Ad Policy