Twitter Files
≈ 19 min read
Names in other languages
No names in other languages are held for this entry yet.
Article bodies are English and are not machine-translated. A name in another language is not a translation of the article.
The Twitter Files are a series of releases of select internal Twitter, Inc. documents published from December 20221 through March 2023 on Twitter. CEO Elon Musk gave the documents to journalists Matt Taibbi, Bari Weiss, Lee Fang, and authors Michael Shellenberger, David Zweig, Alex Berenson, and Paul D. Thacker shortly after he acquired Twitter on October 27, 2022. Taibbi and Weiss coordinated the publication of the documents with Musk, releasing details of the files as a series of Twitter threads. After the first set of files was published, various technology and media journalists said that the reported evidence demonstrated little more than Twitter's policy team struggling with difficult decisions, but resolving such matters swiftly. Some conservatives said that the documents demonstrated what they called Twitter's liberal bias. A major aspect of the examination surrounded false assertions by Musk and others that Twitter had been ordered by the government to help presidential candidate Joe Biden in the coming election by suppressing an October 2020 New York Post story about Hunter Biden's laptop. Researcher Matt Taibbi found no evidence of government involvement in Twitter's decision to initially withhold the story.…
The Twitter Files are a series of releases of select internal Twitter, Inc. documents published from December 2022 through March 2023 on Twitter. CEO Elon Musk gave the documents to journalists Matt Taibbi, Bari Weiss, Lee Fang, and authors Michael Shellenberger, David Zweig, Alex Berenson, and Paul D. Thacker shortly after he acquired Twitter on October 27, 2022. Taibbi and Weiss coordinated the publication of the documents with Musk, releasing details of the files as a series of Twitter threads. After the first set of files was published, various technology and media journalists said that the reported evidence demonstrated little more than Twitter's policy team struggling with difficult decisions, but resolving such matters swiftly. Some conservatives said that the documents demonstrated what they called Twitter's liberal bias. A major aspect of the examination surrounded false assertions by Musk and others that Twitter had been ordered by the government to help presidential candidate Joe Biden in the coming election by suppressing an October 2020 New York Post story about Hunter Biden's laptop. Researcher Matt Taibbi found no evidence of government involvement in Twitter's decision to initially withhold the story. In a June 2023 court filing, Twitter attorneys strongly denied that the Files showed the government had coerced the company to censor content, as Musk and many Republicans claimed. Former Twitter employees asserted that Republican officials also made takedown requests so often that Twitter had to keep a database tracking them. Internal Twitter emails showed the company allowed accounts operated by the U.S. military to run a Middle East influence campaign; some accounts were kept on the platform for years before being taken down. The releases prompted debate over the nature of blacklisting, vows for congressional investigation, calls for the full release of all documents for the sake of transparency, and calls to improve content moderation processes at Twitter.
Background
The inner workings of Twitter's content moderation systems were not well known to the public, on the basis that knowledge of the details could enable manipulation. But American conservatives had long contended that Twitter used its moderation policies to muzzle conservative views. On November 28, 2022, a month after Musk officially acquired Twitter, Musk announced that he planned to release a portion of Twitter's internal documents related to "free speech suppression", adding, "The public deserves to know what really happened" under Twitter's prior leadership. Musk subsequently gave a series of internal Twitter documents—including screenshots, emails, and chat logs—to freelance journalists Matt Taibbi and Bari Weiss. Taibbi noted that "in exchange for the opportunity to cover a unique and explosive story, I had to agree to certain conditions" that he did not disclose. Weiss stated that the only condition she and her reporting team agreed to was that the material would be first published on Twitter. Musk later stated he had not read the documents before their release to Taibbi and Weiss. On December 6, Musk fired James Baker, deputy general counsel at Twitter, for allegedly vetting the information before it was passed on to Taibbi and Weiss and providing an explanation that Musk found "unconvincing." Taibbi said that the planned publication of Twitter's internal documents related to its handling of the Hunter Biden laptop story had been delayed because of Baker's vetting. Baker had previously been general counsel for the FBI and investigated Russian interference into the 2016 election.
Topics
In his prelude, Taibbi stated that the files told a "Frankenstein tale of a human-built mechanism"—"one of the world's largest and most influential social media platforms"—"grown out [of] the control of its designer". Taibbi wrote that these documents, as well as the assessment of "multiple current and former high-level executives", demonstrate how, although external requests for moderation from both political parties were received and honored, an overwhelmingly left-wing employee base at Twitter facilitated a left-leaning bias. The first installment included content related to Twitter's moderation process regarding a New York Post article on the Hunter Biden laptop story. The second installment addressed what Musk and others have described as the shadow banning of some users. The third installment highlighted events within Twitter leading to President Donald Trump's suspension from Twitter. The fourth installment covered how Twitter employees reacted to the January 6 United States Capitol attack and the conflict within Twitter on how to moderate tweets and users supporting the attack. The fifth installment covered how Twitter employees influenced the decision to ban Trump from the platform. The sixth installment described how the FBI contacted Twitter to suggest that action be taken against several accounts for allegedly spreading election disinformation. The seventh installment showed Twitter's interaction with the intelligence community around the New York Post story on Hunter Biden's laptop. The eighth installment showed the Twitter Site Integrity Team whitelisted accounts from United States Central Command (CENTCOM) used to run online influence campaigns in other countries.
No. 1: Content moderation of New York Post story
, who published the first installment of the documents On December 2, 2022, Taibbi posted a lengthy Twitter thread reporting on the first installment of the Twitter Files, which he illustrated with images of some of the files. Taibbi's installment attracted thousands of retweets. Some documents described Twitter's internal deliberations regarding the decision to moderate content relating to the Hunter Biden laptop controversy, while others contained information on how Twitter treated tweets that were flagged for removal at the request of the 2020 Biden campaign team and the first Trump White House. He also shared communications between California Democrat Ro Khanna and then-Twitter head of legal Vijaya Gadde, in which Khanna warned about the free-speech implications and possible political backlash that would result from censorship. The laptop controversy related to a 2020 New York Post article that presented allegations concerning a laptop computer of Hunter Biden, son of then-presidential candidate Joe Biden. Twitter, along with Facebook, implemented measures to block its users from sharing links to the story, and Twitter further imposed a temporary lock on the accounts of the New York Post and White House Press Secretary Kayleigh McEnany, citing violations of its rules against posting hacked content. The Washington Post reported that this was a result of the company's scenario-planning exercises to combat disinformation campaigns, which included potential "hack and leak" situations like what had transpired during the Russian interference in the 2016 United States elections. The decision generated an outcry from then-President Trump and conservatives who saw it as politically motivated. Yoel Roth, then Twitter's Head of Trust and Safety, later said he had not been in favor of withholding the story and acknowledged that it was a "mistake" to censor it. The installment shed light on an internal debate on whether Twitter should prevent the story from being shared, with leadership arguing that it fell under the company's prohibition on hacked materials. According to Taibbi, then-CEO Dorsey was unaware of the decision to suppress the content when it was made. Days later, Dorsey reversed the decision, calling it a "mistake", and Twitter updated its hacked materials policy to state that news stories about hacked materials would be permitted, but with a contextual warning. Taibbi also shared a screenshot of what appeared to be a request from the Biden campaign asking for a review of five tweets, along with the Twitter moderation team's reply, "Handled these." Taibbi did not disclose the content of those tweets, but four were later found from internet archives to contain nude images of Hunter Biden, which violated Twitter policy and California law as revenge porn; the content of the fifth deleted tweet is unknown. Musk tweeted that Twitter had acted "under orders from the government", though Taibbi reported that he found no evidence of government involvement in the laptop story, tweeting, "Although several sources recalled hearing about a 'general' warning from federal law enforcement that summer about possible foreign hacks, there's no evidence—that I've seen—of any government involvement in the laptop story." His reporting seemed to undermine a key narrative promoted by Musk and Republicans that the FBI pressured social media companies to suppress the Hunter Biden laptop stories.
No. 2: Visibility filtering
Weiss published the second installment on December 8, covering "visibility filtering." Twitter "rank[s]" tweets and search results, promoting some tweets for "timely relevance" and limiting the exposure of others. The company uses the term "visibility filtering" to refer to these practices as well as user-generated filtering—such as when one user blocks or mutes another account. One goal of visibility filtering is to reduce the reach of accounts that violate Twitter rules without committing violations egregious enough to warrant suspension. Weiss contended that "visibility filtering" was merely Twitter's in-house term for "shadow banning". She posted screenshots of employee views of user accounts with tags indicating visibility filtering, and wrote that politically sensitive decisions were made by the Site Integrity Policy, Policy Escalation Support (SIP-PES) team, which included the chief legal officer, head of trust and safety, and CEO. She posted screenshots of the accounts of Stanford professor Jay Bhattacharya (an opponent of COVID-19 lockdowns), conservative radio host Dan Bongino, and conservative activist Charlie Kirk, which were respectively tagged with "Trends Blacklist", "Search Blacklist", and "Do Not Amplify". She also said that the SIP-PES team was responsible for the multiple suspensions of the anti-LGBT account Libs of TikTok, which had been tagged with "Do Not Take Action on User Without Consulting With SIP-PES". She noted that Twitter had not taken down a tweet containing the address of the account's owner, Chaya Raichik. Weiss characterized these practices as censorship and as evidence of shadow banning, which Twitter disputed, largely on the basis of its different definition of "shadow ban". Twitter distinguished visibility filtering from shadow banning, which it defined as making "content undiscoverable to everyone except the person who posted it." The documents Weiss discussed focused on individuals popular with the right-wing and suggested the moderation practices were politically motivated —a long-standing claim among American conservatives, which Twitter has denied. An internal study Twitter conducted in 2018 found its algorithms favored the political right. Wired and Slate described the policy by which moderators were unable to act on high-profile conservative accounts without first escalating to high-level management as "preferential treatment", since this effectively limited Twitter's enforcement of their content policies on these accounts. Weiss did not reveal how many accounts overall were de-amplified nor the politics of those who were, and this lack of context made it difficult to glean any conclusions on the matter. Kayvon Beykpour, the former head of product at Twitter, called the installment "deliberately misleading"; in the interest of transparency, Dorsey called for all of the Twitter Files to be released, tweeting to Musk, "Make everything public now."
Nos. 3–5: Attack on the Capitol and suspension of Donald Trump
The third installment was published by Taibbi on December 9, highlighting the events within the company that led up to Trump's suspension from Twitter. Two days after the January 6, 2021, United States Capitol attack, Trump made two tweets: one praised his voters, calling them "American Patriots" who will "not be disrespected or treated unfairly in any way, shape or form!!!" and the other stated that he would not be attending Joe Biden's inauguration. Twitter permanently suspended Trump's account on the same day, citing the two tweets as a violation of the "glorification of violence" policy. Taibbi reported that on October 8, 2020, Twitter executives created a channel entitled "us2020_xfn_enforcement" as a hub to discuss content removal that pertained to the then-upcoming 2020 United States presidential election. Twitter's moderation process was, according to Taibbi, based on guesswork, "gut calls", and Google searches, including moderation of then-President Trump's tweets. As previously reported by The New York Times in 2020, Taibbi said that then-head of Trust and Safety for Twitter, Yoel Roth, met regularly with agencies such as the FBI to discuss potential attempts by foreign and domestic actors to manipulate the 2020 election. Following the suspension of Trump's Twitter account, Taibbi reported that it set a precedent for the suspension of future presidents' accounts, which he claimed to violate Twitter's own policies. Taibbi wrote that he was told that the Trump administration and Republicans had made requests to moderate tweets but did not find any evidence of these requests in the election enforcement Slack chat. The fourth installment was published on December 10 by Shellenberger. It covered how Twitter employees reacted to the January 6 United States Capitol attack and the conflict within the company about how to take action against tweets and Twitter users who were supporting the attack without a specific policy as backing, due to the unprecedented nature of Trump's false claims of winning the 2020 United States presidential election. Shellenberger shared screenshots of Roth asking a coworker to blacklist the terms "stopthesteal" and "kraken", both of which were associated with supporters of the January 6 attack. He also said that pressure from the company's employees appeared to influence Dorsey to approve a "repeat offender" policy for permanent suspension. After receiving five strikes as per the new policy, Trump's personal Twitter account was permanently suspended on January 8. Shellenberger's installment also provided screenshots suggesting that there were instances when employees flagged tweets and applied strikes at their own discretion without specific policy guidance, which according to Shellenberger, are examples of a frequent occurrence. The fifth installment was published on December 12 by Weiss. It covered the conflict between Twitter employees and how it influenced the decision regarding Trump's ban from the platform. Those communications include requests from the FBI and other agencies to determine if a particular tweet violated policies against election manipulation. Weiss reported that two tweets Trump made on the morning of January 8, 2021, were used as a foundation for his suspension. She said that the two tweets were initially cleared as having no indication of incitement of violence, to the agreement of multiple employees. According to Weiss, former head of Legal, Policy, and Trust Vijaya Gadde dissented, suggesting that the tweets were dog whistles for future political violence. Weiss reported that Twitter's "scaled enforcement" team engaged and agreed with Gadde, suggesting that the tweets violated the "glorification of violence" policy and that the term "American Patriots" Trump used in a tweet was code for the Capitol rioters. She also said that one team member referred to Trump as a "leader of a terrorist group responsible for violence/deaths comparable to the Christchurch shooter or Hitler". Weiss reported that after a 30-minute all-staffer meeting, Dorsey asked Roth to simplify the language of the document for Trump's suspension. One hour later, Trump's account was suspended "due to the risk of further incitement of violence".
Nos. 6–7: FBI communications with Twitter Trust and Safety Team
The sixth installment, published by Taibbi on December 16, described how the FBI reported several accounts to Twitter's Trust and Safety Team for allegedly spreading election misinformation. According to Taibbi, many of the accounts reported had small numbers of followers and were making tweets seemingly satirical in nature, such as user Claire Foster who had tweeted "I'm a ballot counter in my state. If you're not wearing a mask, I'm not counting your vote. #safetyfirst" and "For every negative comment on this post I'm adding another vote for the democrats". He also reported that Twitter did not always take action against tweets and accounts flagged by the FBI. Taibbi wrote that a high-ranking staff member referred to the company's relationship with the FBI as "government-industry sync" due to the frequency of emails and meetings with the agency. The seventh installment was published by Shellenberger on December 19, 2022. It described the FBI's involvement in moderating the Hunter Biden laptop story. Shellenberger reported that the FBI's and the DHS' warnings about potential foreign interference in the 2020 presidential election influenced Twitter to moderate the Hunter Biden laptop story. Roth wrote in an internal discussion about the Post story that due to "the SEVERE risks here and lessons of 2016", Twitter should apply a warning to the story and prevent it from "being amplified". Shellenberger shared screenshots of an email from 2021, which included a communication from Twitter's Safety, Content, & Law Enforcement (SCALE) team that Twitter had received $3,415,323 from a 2019 program designed to meet the "statutory right of reimbursement" for the cost of processing requests from the FBI. Musk claimed in a tweet that this payment is proof of the U.S. government bribing the company "to censor info from the public", despite such payments being commonplace for processing legal requests. Twitter's guidelines under law enforcement state that "Twitter may seek reimbursement for costs associated with information produced pursuant to legal process and as permitted by law (e.g., under 18 U.S.C. §2706)". Alex Stamos, former chief security officer at Facebook and partner at cyber consulting firm Krebs Stamos Group, wrote that the reimbursements from the FBI have "absolutely nothing to do with content moderation".
Congressional committee investigations
Republican House members Jim Jordan and James Comer launched complementary investigations in January 2023. Former Twitter employees and members of the Twitter Files team provided testimony. Both committees concluded their investigations in 2024.
Wikipedia article
Musk accused Wikipedia of "non-trivial left-wing bias" after the Twitter Files article was considered for deletion, replying to screenshots of select users referring to it as "not notable" and a "nothing burger"; however, the final decision was to keep the article.
Missouri v. Biden
On March 24, 2026, the Trump administration reached a settlement agreement in Missouri v. Biden, promising that three federal agencies would no longer pressure companies to remove or suppress content. This resolves the lawsuit brought by Missouri and Louisiana and other plaintiffs that the Biden administration had threatened social media platforms' posts about the 2020 election, the Hunter Biden laptop controversy and COVID-19. The Twitter Files and congressional investigations showed extensive government–platform coordination on content moderation.
References
External links
Capsule Summaries of all Twitter Files Threads to Date, With Links and a Glossary - Files hosted on Matt Taibbi's Racket News at Substack. "Hearing on Twitter Documents About Content Moderation Decisions." U.S. House Judiciary Select Subcommittee on the Weaponization of the Federal Government. C-SPAN. March 9, 2023.
Text derived from https://en.wikipedia.org/w/index.php?oldid=1375616867 under CC BY-SA 4.0; modifications have been made.
References
2 claims could not be tied to a specific sentence and are listed in the table below instead.
- ↩Wikipedia “Twitter Files” publication date (×1) · confidence 0.95 revision URL