How the FBI Hacked Twitter
The answer begins with Russiagate
After journalist Matt Taibbi published the first batch of internal Twitter documents known as the Twitter files, he tweeted that the company’s deputy general counsel, James Baker, was vetting them.
“The news that Baker was reviewing the ‘Twitter files’ surprised everyone involved,” Taibbi wrote. That apparently included even Twitter’s new boss, Elon Musk, who added that Baker may have deleted some of the files he was supposed to be reviewing.
Baker had been the top lawyer at the FBI when it interfered in the 2016 presidential election. News that he might have been burying evidence of the spy service’s use of a social media company to interfere with the 2020 election, is rightly setting off alarm bells.
In fact, the FBI’s penetration of Twitter constituted just one part of a much larger intelligence operation—one in which the bureau offshored the machinery it used to interfere in the 2016 election and embedded it within the private sector. The resulting behemoth, still being built today, is a public-private consortium made up of U.S. intelligence agencies, Big Tech companies, civil society institutions, and major media organizations that has become the world’s most powerful spy service—one that was powerful enough to disappear the former president of the United States from public life, and that is now powerful enough to do the same or worse to anyone else it chooses.
Records from the Twitter files show that the FBI paid Twitter nearly $3.5 million, apparently for actions in connection with the 2020 election and
nominally a payout for the platform’s work censoring “dangerous” content that had been flagged as mis- or disinformation. That “dangerous” content notably included material that threatened Joe Biden and implicated U.S. officials who have been curating the Biden family’s foreign corruption for decades.
The Twitter files have to date focused on FBI and, to a lesser extent, CIA election interference. However a lesser-known U.S. government agency, the Cybersecurity and Infrastructure Security Agency (CISA) also played a significant role in shaping the 2020 vote. “CISA is a sub-agency at DHS that was set up to protect real physical infrastructure, like servers, malware and hacking threats,” said former State Department official Mike Benz, now the executive director of the Foundation for Freedom Online. “But they expanded ‘infrastructure’ to mean us, the U.S. electorate. So ‘disinformation’ threatened infrastructure and that’s how cybersecurity became cyber-censorship. CISA’s mandate went from stopping threats of Russian malware to stopping tweets from accounts that questioned the integrity of mail-in voting.”
We have some insight into CISA’s de facto censorship of Twitter because their private-sector partners boasted about such activities in promotional material. One such public-private partnership was the Election Integrity Partnership (EIP), a censorship consortium consisting of the Stanford Internet Observatory, the University of Washington’s Center for an Informed Public, the Atlantic Council’s Digital Forensic Research Lab, and Graphika, a D.C.-based private company founded by former national security officials. According to a document from the Twitter files release, Graphika is employed by the Senate Intelligence Committee for “narrative analysis and investigations.” For CISA, Graphika and its EIP partners served as an intermediary to censor social media during the 2020 election cycle.
CISA targeted posts questioning the election procedures introduced into the election process on account of COVID-19, like mass mail-in ballots, early voting drop boxes, and lack of voter ID requirements. But instead of going to the platforms directly, CISA filed tickets with EIP, which relayed them to Twitter, Facebook, and other tech companies. In “after-action” reports, the Election Integrity Partnership bragged about censoring Fox News, the New York Post, Breitbart, and other right-leaning publications for social media posts and online links concerning the integrity of the 2020 election.
The censorship industry is based on a “whole of society model,” said Benz. “It unifies the government and the private sector, as well as civil society in the form of academia and NGOs and news organizations, including fact-checking organizations. All these projects with catchphrases like building resilience, media literacy, cognitive security, etc., are all part of a broad partnership to help censor opponents of the Biden administration.”
Notably, Baker was enlisted in one of the civil-society organizations at the same time he joined Twitter as deputy general counsel. According to Benz, the National Task Force on Election Crises is something like a sister organization to the Transition Integrity Project, the group founded by former Democratic Party officials and Never Trump publicists who war-gamed post-2020-election scenarios. “The outfit Baker was part of,” said Benz, “effectively handled the public messaging for an organization that threatened street violence and counseled violating the constitution to thwart a Trump victory.”
Baker’s presence at Twitter, then, and his review of the Twitter files, was deeply disconcerting. “This is who is inside Twitter,” the journalist and filmmaker Mike Cernovich tweeted at Elon Musk this spring. “He facilitated fraud.”
Musk replied: “Sounds pretty bad.”
In fact, Musk has done more in two months to bring to light crimes committed by U.S. officials than William Barr and John Durham did during their three-year investigation of the FBI’s election interference activities during the 2016 election. Musk now owns what became a crucial component of the national security apparatus that, seen in this light, is worth many times more than the $44 billion he paid for it.
The FBI prepared America’s new public-private censorship regime for the 2020 election by falsely telling Twitter, as well as other social media platforms, press outlets, lawmakers, and staff members of the White House, that Russians were readying a hack and leak operation to dirty the Democratic candidate. Accordingly, when reports of a laptop owned by Hunter Biden and giving evidence of his family’s financial ties with foreign officials were published in October 2020, Twitter blocked them.
In the week before the election, the FBI field office in charge of investigating Hunter Biden sent multiple censorship requests to Twitter. The FBI has “some folks in the Baltimore field office and at [FBI headquarters] that are just doing keyword searches for violations,” a company lawyer wrote in a Nov. 3, 2020, email.
The documents also show that Twitter banished Trump after misrepresenting his posts as incitement to violence. With U.S. intelligence services reportedly using informants to provoke violence during the January 6th protest at the Capitol, the trap closed on Trump. Twitter and Facebook then moved to silence the outgoing president by denying him access to the global communications infrastructure.
The FBI unit designated to coordinate with social media companies during the 2020 election cycle was the Foreign Influence Task Force. It was set up in the fall of 2017 “to identify and counteract malign foreign influence operations” through, “strategic engagement with U.S. technology companies.” During the election cycle, according to the Twitter files, the unit “swelled to 80 agents and corresponded with Twitter to identify alleged foreign influence and election tampering of all kinds.”
The FBI’s chief liaison with Twitter was Elvis Chan, an agent from its Cyber Branch. Based in the San Francisco field office, Chan was also in communication with Facebook, Google, Yahoo!, Reddit and LinkedIn. Chan demanded user information that Twitter said it could not release outside of a “legal process.” In exchange, Chan promised to secure temporary security clearances for 30 Twitter employees a month before the election, presumably to give staff the same briefings on alleged Russian information operations provided to U.S. officials in classified settings.
But Twitter executives claimed they found little evidence of Russian activity on the site. So Chan badgered former head of site security Yoel Roth to produce evidence the FBI was serving its advertised mission of combating foreign influence operations when in fact it was focused on violating the First Amendment rights of Americans.
Chan briefed Twitter extensively on an alleged Russian hacking unit, APT28, or Fancy Bear, which was the same outfit that was claimed by Hillary Clinton campaign contractors to have hacked and leaked Democratic National Committee emails in 2016. According to Roth, the FBI had “primed” him to attribute reports about Hunter Biden’s laptop to an APT28 hack-and-leak operation. Needless to say, the FBI’s reports—and subsequent “disinformation” claims—were themselves blatant disinformation, invented by the FBI, which had been in possession of the laptop for nearly a year.
The FBI’s penetration of Twitter constituted just one part of a much larger intelligence operation—one in which the bureau offshored the machinery it used to interfere in the 2016 election and embedded it within the private sector.
Twitter was more than a one-way mirror: The FBI also seems to have embedded its own spy structure within the social media company to siphon off the personal data and behavior of users. Dozens of former intelligence officials were installed within Twitter after the election of Donald Trump. Some had active top secret security clearances. Twitter’s director of strategy was Dawn Burton, former FBI Director James Comey’s deputy chief of staff. Perhaps most significant was Baker himself, who appears to have led the FBI’s internal organization at the platform. Efforts to reach Baker for comment on this story were unsuccessful.
Baker left the FBI in 2018 under a cloud of suspicion. In 2017, the Justice Department investigated him for leaking to the press, and the Republican-led House of Representatives later investigated him for his role in Russiagate. Former congressional officials say that as part of the bureau’s 2016 investigation of the Trump campaign, Baker authored the warrant to spy on Trump’s inner circle.
After he departed the law enforcement agency, CNN rewarded him for his “resistance” activities—which boosted the network’s ratings to record levels—by hiring him as a legal analyst. The Washington, D.C.-based Brookings Institution brought Baker on board to contribute to its collusion-conspiracy website “Lawfare.” DOJ again investigated him in 2019 for leaking to the media while at the FBI. In June 2020, Baker joined Twitter as deputy general counsel. With his security clearances still active, he was Twitter’s liaison with U.S. intelligence agencies, where he reinforced the FBI’s external pressure from inside Twitter to censor the Biden laptop story.
Under Baker, Twitter became more than just an instrument to censor the opposition; it also spied on them. Newly released court documents show that Twitter coordinated with the DOJ to intercept the communications of users potentially dangerous to the Biden campaign, like Tara Reade, the former Biden Senate staffer who alleged that Biden had sexually assaulted her decades earlier. The DOJ subpoenaed her Twitter account, likely with the purpose of giving the company cover for finding out which journalists had contacted her about her allegations.
The cozy two-way relationship between the government and the social media company, which Baker helped oversee and ultimately used to interfere in the 2020 election, was years in the making. In 2014, Twitter filed suit against the DOJ and the FBI, Twitter v. Holder. The San Francisco-based social media platform had been served Foreign Intelligence Surveillance Act (FISA) warrants to collect the electronic communications of some of its users, and Twitter said that in the interests of transparency, it wanted to release a public report with the precise numbers of warrants it had been served. FBI General Counsel James Baker refused. Twitter could disclose the number of warrants in broad, inexact ranges, for instance between 0-249, but not the exact number, even if it was zero.
To press its case, Twitter hired Perkins Coie, a prominent Democratic Party-aligned firm that had represented several presidential campaigns—John Kerry’s, Barack Obama’s and eventually Hillary Clinton’s 2016 run, during which the law firm would hire Fusion GPS to produce the discredited dossier of Trump-Russia reports under the byline of British ex-spy Christopher Steele that became the center of Russiagate.
The firm’s lead attorney for Twitter v. Holder was former DOJ cybersecurity expert Michael Sussmann. He and Baker were friends. The FBI lawyer thanked him in a September 2014 letter for a recent meeting that included Twitter’s top lawyer Vijaya Gadde and others, but affirmed that giving specific numbers would reveal “properly classified information.” Why that would endanger sources and methods, as the government claimed, Baker never explained. But no one at DOJ knew more than Baker about FISA, the most intrusive surveillance program that U.S. intelligence services have in their arsenal.
Even during his time in the private sector Baker had worked on FISA issues. In 2008, he’d taken a job at Verizon, where George H.W. Bush’s former Attorney General William Barr was general counsel. Baker was assistant general counsel for national security, and thus an entry point for his former DOJ colleagues, facilitating their access to material obtained through FISA and other surveillance programs. It wouldn’t have occurred to him or Barr to want to publish, as Twitter said it did, the number of FISA warrants that law enforcement served their private-sector employer. They were DOJ men, and FISAs are highly classified. Few outside the intelligence community had ever seen one, until the Trump era.
An April 2017 story in The Washington Post disclosed that the FBI had obtained a FISA warrant to spy on Trump campaign volunteer, Carter Page, making FISA part of the national lexicon. The Post story, sourced to law enforcement and other U.S. officials, far exceeded what Twitter was prevented from publishing for national security reasons. It named the subject of a FISA warrant, and revealed that the warrant targeted a presidential candidate’s circle.
“Baker authored the Page FISA and signed off on all of it,” Kash Patel, a member of President Trump’s National Security Council, told me. Patel also served as Devin Nunes’ lead investigator for the House Intelligence Committee’s probe of FBI crimes and abuses committed during the bureau’s Trump-Russia investigation. Recently, reports have also surfaced that the DOJ was spying on Patel and other Nunes staffers while they were investigating the FBI and DOJ.
Patel continued. “When I was at DOJ,” he said, “Baker had a reputation of being a FISA guru. The Page FISA was crafted by someone who knew what questions not to ask, and how to use language to get it past a FISA court judge without fully disclosing facts they knew would have disqualified the warrant.”
Baker told congress that he didn’t normally work on FISAs in his job as the FBI’s top lawyer, but this FISA was especially sensitive: It allowed the bureau to sweep up a presidential campaign’s electronic communications, including those of a certain Republican candidate. So, Baker said, he “wanted to make sure that we were filing something that would adhere to the law.”
But the Page FISA was unlawful. The FBI had simply laundered the Clinton campaign’s anti-Trump dirt into a surveillance warrant so it could justify spying on the candidate in support of her Democratic rival. “The FBI wanted the warrant, so they wrote it in a way to get it even though they knew it was a fraud, as our investigation would expose,” said Patel. But with Baker squaring away the package, who was going to question the FISA guru?
By fall 2016, Baker had become the preferred drop box for the Clinton team to push anti-Trump dirt into the FBI. His friend, the journalist David Corn, passed him more Steele reports, which he handed off to FBI colleagues investigating the GOP candidate. Baker also agreed to a meeting with a former associate who wanted to pass on research from cyber experts about a supposed secret computer channel between a Russian bank and Trump. That was Michael Sussmann.
Five years later, the special counsel appointed to investigate the FBI’s Trump-Russia probe would charge Sussmann with lying to the FBI. Specifically, he’d lied to his friend Jim Baker: When Sussmann met with him in September 2016 to pass on Trump-Russia information, he told Baker he wasn’t representing a client when in fact he was working for the Clinton campaign.
Given what we know now, it’s clear that special counsel John Durham’s case against Sussmann was even more troubled than it first seemed. His star witness, Baker, wasn’t a hero in the story but a co-conspirator, to whom Durham gave a pass so he could charge Sussmann with a process crime.
Obviously, Baker knew his friend was representing the Clinton campaign—that’s what Perkins Coie lawyers do: represent Democratic Party presidential campaigns. But the two would want to cover their tracks, so before their September 2016 meeting, the Clinton lawyer sent Baker a text saying that he had information to share, and he wasn’t representing a client. This would have proven Durham’s case about Sussman lying to his friend Baker at the FBI, except Baker never told the prosecutor about the text. Neither did the DOJ’s inspector general, who had Baker’s phones, until it was too late to use the evidence in court.
What few understood was that the issue wasn’t just the 2016 election but the 2020 vote, too. Baker had to tread carefully or else risk exposing the job for which Sussmann had helped plant him at Twitter. It was one of the spy service’s most sensitive operations—infiltrating social media platforms to fix a presidential race. So Sussmann was acquitted—and the FBI’s hack of Twitter continued.
The Twitter files’ disclosures about the coordination between the company and spy agencies to fix presidential elections sheds light on the nature of Twitter v. Holder, which was eventually decided in the government’s favor shortly before Baker joined the company. Twitter filed the suit because it believed in transparency—and to reassure users that the platform wasn’t being used to spy on them, or not most of them. But something else was going on behind the scenes: Social media platforms were already being assimilated into the intelligence services.
Documents leaked in 2013 by former National Security Agency contractor Edward Snowden showed that the NSA was mining social media platforms to build profiles on Americans. Previously, the NSA was required to stop searching the contact chain of a foreign target when it reached a U.S. citizen, but a 2010 change in policy allowed the intelligence services to trace the contacts of Americans so long as there was a “foreign intelligence” purpose. That is, even at the dawn of the social media revolution, the spy services saw social media as a surveillance tool, like FISA.
In response to Snowden’s disclosures, then-President Barack Obama gave high-minded speeches about balancing civil rights and national security. But by the time Twitter filed its 2014 suit, the White House had already chosen to turn surveillance programs against its domestic opponents. Obama’s intelligence chiefs spied on U.S. legislators and pro-Israel activists opposed to Obama’s signature foreign policy initiative, the Iran nuclear deal.
The Obama administration also realized that it could lean hard on monopoly social media platforms in order to gain political advantages—and it could make companies that weren’t compliant pay a price. First strike got you a dressing down from the White House: Weeks after the 2016 vote, for instance, Obama pulled Mark Zuckerberg aside at a conference in Peru and read him out about not doing more to keep Russian disinformation off Facebook. The reality is that Russia spent around $135,000 on Facebook ads, a small percentage of what presidential campaigns typically spend on a single day before lunch. But Obama wasn’t worried about Russia—he struck deals with Vladimir Putin to advance his own idiosyncratic foreign policy goals, like the nuclear agreement with Russia’s ally Iran. Obama’s problem was Trump.
As he was leaving office, Obama stamped the U.S. government’s seal of approval on Russiagate, ordering his spy chiefs to draft an official assessment claiming Putin helped put Trump in the White House. Since then, in Deep State parlance, “Russia” equals Trump and stopping “Russian disinformation” means censoring Trump, his supporters, and anyone else opposed to the national security apparatus’s takeover of the public communications infrastructure. Since Zuckerberg didn’t keep Trump off Facebook in 2016, he had to put up $400 million to drive votes to Democrats in 2020—and even that wasn’t enough. In 2021, Democratic Party insiders working together with Zuckerberg’s Big Tech competitor, eBay founder Pierre Omidyar, sent a fake whistleblower after him to testify before congress that Facebook was bad for teenage girls.
The censorship regime would regulate out of existence anyone who resists it. To make the case for the hegemony of government censors, it found an eminent pitchman: Barack Obama.
In April, as Musk first said he wanted to buy Twitter and save free speech, Obama embarked on a “disinformation” tour, which took him to several college campuses to promote the un-American virtues of censorship. He first visited his hometown to speak at a University of Chicago conference, “Disinformation and the Erosion of Democracy.” Other guests included Anne Applebaum, an early advocate of the collusion conspiracy theory who pushed the spy-service fiction in dozens of her Washington Post columns. Also in attendance was former CISA head Chris Krebs, now famous for congressional testimony in which he claimed the 2020 election was the most secure ever.
EIP principals from the Stanford Internet Observatory were featured speakers at the daylong seminar at the Palo Alto university where Obama made the second stop on his April “disinformation” tour. Regulation, Obama told the Stanford audience, has to be part of the answer to solve the disinformation crisis. In other words, he went to Silicon Valley to threaten his listeners that he would ruin their financial model by stripping away social media’s liability exemptions.
The purpose of Obama’s speech was to present a choice to his audience: Either you impose a scorched-earth policy against the establishment’s opponents, or else you will face the kind of regulation that every company knows will be its death knell. Moreover, if they made the right choice, Obama showed, there was money in it for them.
“In effect, Obama announced that the funding channels are open for people who want to do disinformation work,” said Mike Benz, executive director of the Foundation for Freedom Online. “It’s like what happened with climate change. If you were an academic who wanted federal funding for anything, you made sure you made reference to climate to get grants. Same now with disinformation. Obama was saying, ‘here’s where the puck is moving, so skate here if you want federal funding.’”
To reward EIP for greasing its path to the White House, the Biden administration awarded all four consortium partners with grant money. The Stanford and Washington units received $3 million from the National Science Foundation “to study ways to apply collaborative, rapid-response research to mitigate online disinformation.” Graphika won nearly $5 million from the Pentagon after the 2020 election for “research on cross-platform detection to counter malign influence,” and another $2 million in 2021. Since 2021, the Atlantic Council received $4.7 million in federal grants, mostly from the State Department, a total far exceeding its previous awards.
In retrospect, the failure of the Russiagate conspiracy theory accelerated the spy service’s takeover of social media. Though no one is likely to be held accountable anytime soon, or ever, it was enough that the details of the operation were exposed by Patel and Nunes. In response, the spy agencies moved much of their operations out of the federal government and into the private sector, where even if congressional investigators found it, there wasn’t much they could do. Republicans could threaten to regulate social media, but their threats were empty. They might even find themselves—and their campaign ads—banned from Twitter.
The public-private sector merger worked only because, as a unifying myth for the U.S. elite, if not as a legal or political maneuver, Russiagate was a great success. If there were any fears of how news of the FBI’s spying operation on a presidential campaign might be received by the press, civil rights activists, and the left, the reception to Russiagate dispelled those concerns. The media offered itself up as a platform for information operations and published illegal leaks of classified information while the rest of the ruling class promoted a conspiracy theory and celebrated the assault on the constitutional rights of their fellow Americans as a success story.
The Republican attorney general of the United States, William Barr—the ultimate DOJ insider—knew the FBI was working to fix the 2020 election and did nothing to stop it. His Justice Department had the laptop in its possession and Barr knew it was authentic. He told reporters this spring that he was “shocked” Biden lied about his son’s computer in the Oct. 22, 2020, debate with Trump. “He’s squarely confronted with the laptop, and he suggested that it was Russian disinformation,” said Barr, “which he knew was a lie.” Yet agents under Barr’s authority were briefing that lie to social media platforms, the press, Congress, and even the Trump White House.
“There were 80 FBI agents in the unit working on foreign disinformation,” Patel told me. “It was about a presidential election, so it would require authorization from the FBI director and the attorney general. Barr knew.”
Barr resigned from the administration a month after the election, outraged that Trump kept pushing him to investigate election fraud when, according to Barr, there was no evidence of it. And yet on his watch, law enforcement agencies under his authority ran the biggest election interference operation in U.S. history. William Barr did not respond to a request for comment.
It seems Barr’s contempt for the president he served blinded him—along with the class of people to which he belongs, Democrats and Republicans alike—to an essential fact: A whole-of-society industry designed to shape elections and censor, propagandize, and spy on Americans was never simply a weapon to harm Donald Trump. It was designed to replace the republic.
