Internal Files Point to Twitter Meddling in 2020 Elections

Then-President Donald Trump looks on during a rally in Washington, on Jan. 6, 2021. Jim Bourg/File Photo/Reuters

The latest episode of the Elon Musk-endorsed “Twitter Files,” internal communications about the social media giant’s “free speech suppression“ on the platform, reveals a soft-on-Biden tough-on-Trump bias in content enforcement, raising questions about how heavily the social media platform put its finger on the scales in the 2020 election.

The latest set of internal Twitter communications, titled “The Removal of Donald Trump,” dives into the actions of Twitter executives during the period from October 2020 to Jan. 8, 2021, when Trump was banned from the platform.

Internal Slack chats at Twitter, shared and commented on by investigative journalist Matt Taibbi in a lengthy thread published on Dec. 9, show that engagement between Twitter executives and federal law enforcement and intelligence agencies soared during this period.

The chats are also chock full of obscure terms and censorious jargon (which Taibbi dubbed “Orwellian unwords”) used by Twitter’s enforcers as they worked to tag, shadow ban, and otherwise suppress content that was sympathetic to then-President Donald Trump’s reelection campaign while giving content aligned with then-candidate Joe Biden’s election efforts a free pass.

The Twitter enforcement team also cracked down on some prominent conservatives weighing in on elections, like former Republican Arkansas Gov. Mike Huckabee and actor James Woods.

When Huckabee posted a joke about mail-in ballots that sparked a debate among Twitter enforcers, Twitter’s former Head of Trust and Safety Yoel Roth said in a Slack channel that he agrees “it’s a joke” but added that Huckabee is “also literally admitting in a tweet to a crime.”

In another exchange, when Twitter staffers didn’t have a “firm policy basis” for censoring a pro-Trump tweet from Woods, a vocally conservative Hollywood actor, they vowed to “hit him hard on future vio with firmer basis.”

“Vio” would be one of Taibbi’s “Orwellian unwords” for “violation.”

Tough on Trump

The messages that comprise episode three of the Twitter Files show how Twitter’s content moderation team came up with various excuses for escalating censorious acts targeting pro-Trump posts.

“In the docs, execs often expand criteria to subjective issues like intent (yes, a video is authentic, but why was it shown?), orientation (was a banned tweet shown to condemn, or support?), or reception (did a joke cause “confusion”?). This reflex will become key in J6,” Taibbi wrote in one of his comments.

One screengrab of Slack messages between Twitter staffers considers action on a tweet by Trump sharing a news report about a postal screw-up where nearly 50,000 voters received incorrect ballots in Ohio, with the news outlet captioning its story with the question whether voters “feel safe mailing in your ballot.”

Trump commented: “No. A Rigged Election!”

A Twitter staffer flagged Trump’s tweet as a “candidate” for getting tagged with the label “Learn how voting by mail is safe and secure.”

Another Twitter moderator questioned whether this would be appropriate as what Trump wrote was “commentary” and “his opinion related to these real events” and so not a violation that justifies a label.

“Yeah... those are factually accurate...” Roth admitted while weighing in on the discussion.

In another set of messages, senior Twitter executives targeted a tweet by Trump that said: “Big problems and discrepancies with Mail In Ballots all over the USA. Must have final total on November 3rd.”

Trump’s tweet was hit with three enforcement actions—a “stay informed” label that invited users to see “how voting by mail is safe and secure,” a tag that said “some or all of the content shared in this Tweet is disputed and might be misleading,” and preventing the tweet from being replied to, shared, or liked.

One of the executives then expressed satisfaction that Trump’s tweet was censored quickly.

“Very well done on speed folks, what this is all designed for,” the executive wrote.

The messages also show that Twitter created a new tool to censor Trump after the election when he was vocal with his claims of election fraud. Internally, executives referred to the tool as “L3 deamplication.”

The new tool was announced on Dec. 10, 2020, when “Trump was in the middle of firing off 25 tweets saying things like, ‘A coup is taking place in front of our eyes,'” Taibbi wrote.

The Twitter enforcement team’s messages showed they sometimes adjusted their moderation actions when the targeted tweets were critical of the censorship itself.

In one instance, Rep. Jody Hice (R-Ga.) posted a tweet sharing a message that had been blocked when shared by another account—“Mailed ballots are more prone to fraud than in-person voting. That shouldn’t be controversial. It’s just common sense.”

Sharing that message, Hice said that Twitter “doesn’t want you to see this tweet” and added, “Say NO to Big Tech censorship.”

One of the Twitter staffers considering action on Hice’s post acknowledges that his message about there being more fraud with mail-in ballots than with in-person ballots “is much more of a legitimate statement even if scale is minuscule” and suggests applying a “soft intervention.”

Roth replies, “Agree” before cautioning that going “too far down the rabbit hole of labeling critical speech (i.e. critical of our handling of this case) is dangerous.”

“It becomes a self-reinforcing ‘wah wah censorship’ cycle,” Roth added.

Soft on Biden

By contrast, there were multiple instances in which pro-Biden tweets warning that Trump “may try to steal the election” were flagged for possible action, only for Twitter executives to give them the green light.

In one such case, a user shared a news report about Supreme Court Justice John Roberts swearing in newly-appointed Justice Amy Coney Barrett, claiming in the comment that “they’re going to try to steal the election” and urging people “if you haven’t voted yet - don’t mail. drop it off or vote early.”

A Twitter staffer asked the team to weigh in, calling it an “edge case” with commentary encouraging voters not to vote by mail and that “I believe we should label it.”

An executive dismissed the recommendation, saying the tweet is “still encouraging people to vote, but expresses the concern that mailed ballots might not make it in time,” which “seems fair.”

In another case, a tweet claiming Trump and Barrett would steal the election was flagged for possible labeling. A Twitter executive decided not to, saying the employee’s concern was “understandable” but the tweet seemed to be in reference to a Supreme Court decision on processing mail-in ballots that arrive after Election Day.

Roth also ordered a label to be overturned that had been applied to a tweet from former Attorney General Eric Holder, who claimed that the U.S. Postal Service had been “deliberately crippled.”

“Everything in it is factually accurate,” Roth wrote, with a Twitter staffer then announcing the label had been removed.

Also, screengrabs of internal messages revealed that after Twitter executives banned Trump from the platform, they discussed banning future presidents, with one executive noting that then-incoming President Joe Biden’s administration “will not be suspended by Twitter unless absolutely necessary.”

‘Federal Election Commission Violations’?

Commenting on the latest Twitter Files revelations, former House Speaker Newt Gingrich told The Epoch Times in an interview that the platform’s behind-scenes operation to suppress certain viewpoints is “pretty horrifying” and shows Twitter was part of the effort to interfere in the 2020 presidential election.

“If you take the president of the United States off of Twitter, what did that cost? If you are Google, and in the last four days of every month, you refuse to deliver Republican fundraising emails, what does that cost?” Gingrich said, referring to Trump’s Twitter ban and allegations that Google had filtered millions of the Republican National Committee’s political campaign emails to spam folders at the end of each month.

“These things should be considered federal election commission violations, because they are actions by a corporation to reshape the election,” Gingrich argued.

“It’s a pretty amazing story of censorship in a country which is dedicated to the First Amendment right of free speech,” he said.

Gingrich believes that major social media platforms like Facebook and Twitter have become so integrated with people’s lives that they should be considered public utilities, and regulated as such.

The Epoch Times has reached out to Twitter for comment.

Caden Pearson and Eva Fu contributed to this report