Elon Musk’s Twitter is beginning to take shape.
A “general amnesty” has restored hundreds of accounts of right-wing activists and QAnon adherents, according to data reviewed by NBC News. The reinstatement of far-right accounts has coincided with a series of bans of left-wing accounts, leaving users unsure of how the company is now applying its rules.
“The ambiguity is a problem,” said Yoel Roth, who recently departed as the head of trust and safety at Twitter. “People don’t know if the rules have changed, and amnesty suggests that at least at some level, there’s disagreement with removing some of those folks. So, we’re seeing a lot of lines being pushed, coupled with more constrained enforcement. It’s a dangerous combination.”
The reinstatements and bans come as researchers continue to monitor an uptick in hate speech, and high-profile users leave the platform. Together, they have created a shift in the platform that has been noticed by Musk’s critics, as well as his supporters.
Musk, who took control of Twitter in late October, has maintained that the company has not changed any of its moderation policies, though Twitter did announce this week that it is no longer enforcing its Covid misinformation policy.
But Musk has also used informal Twitter polls to make major decisions, first to restore the account of former President Donald Trump, and then to enact “general amnesty” to suspended accounts.
“The people have spoken,” he tweeted last week. “Amnesty begins next week.”
Musk kept his word. Travis Brown, an independent software developer in Berlin who tracks Twitter suspensions and screen name changes as part of a project studying extremism, shared a dataset for this article that showed a wide variety of far-right accounts had been reinstated since Musk’s announcement.
In that time, Brown has logged an estimated 12,000 reversals of past bans, in a set that, while not a definitive list of reversals, provides a window into the types of users being welcomed back to the platform and leaving experts alarmed. Among the spammers, copyright rule-breakers, adult-content creators and high-profile accounts, Twitter has reopened the door to a growing and emboldened community of trolls, white nationalists, conspiracy theorists and extreme right-wing activists.
Patrick Casey, a white nationalist, and Andrew Anglin, a neo-Nazi, have both had their accounts restored.
“I never thought I’d see the day when I’d be allowed back on Twitter and yet here we are,” Casey told listeners of his podcast Tuesday.
Those restorations coincide with some users saying they’ve experienced an uptick in harassment that has pushed them to leave.
“Since the takeover, I’ve been getting an increased amount of casual racism and sexual harassment,” Jane Manchun Wong, an independent app researcher based in Hong Kong, who is well-known for breaking insider secrets about features in progress, said Wednesday in a post on the open-source social network Mastodon, citing “the amount of trolls scaling up and empowered on the site.”
Author Sam Harris deleted his account with 1.5 million followers last week, after Trump’s reinstatement. In one of his final tweets, Harris wrote, seemingly referring to Musk, “The prevailing opinion among ‘free speech absolutists’ appears to be that this platform, in order to become healthy, must helplessly publish the malicious lies of any maniac, at scale, regardless of the consequences.”
Many of the accounts in Brown’s set had markers and hashtags in their bios indicating they were engaged in the QAnon conspiracy theory movement, which was largely purged from Twitter last year.
Those reinstatements also come after Twitter dramatically reduced its staff, including those dedicated to dealing with abuse and hate speech. In an interview Wednesday with technology journalist Kara Swisher, Roth expressed doubts about the platform’s ability to enforce its dwindling policies.
“Are there enough people who understand the emergent malicious campaigns that happen on the service and understand it well enough to guide product strategy and policy direction?” he asked. “I don’t think that there are enough people left at the company who can do that work.”
While a dearth of effective content moderation has reportedly turned off advertisers and alienated some users, the move to bring back hoards of former rule-breakers has also drawn attention to the platform, a metric that Musk seems to value, said Sarah T. Roberts, an associate professor at UCLA and the author of “Behind the Screen: Content Moderation in the Shadows of Social Media.” She briefly worked at Twitter researching content moderation, but returned to the school this year.
She called the restoration of banned accounts “a field day” for their holders.
“From their point of view, there’s a sense of vindication and that all you have to do is to wait for the weather to turn,” Roberts said.
That turn also now looks to include new suspensions, in addition to the account restorations.
Several independent media accounts that report on far-right groups and left-wing activists have been suspended in recent days.
Chad Loder is an independent journalist in Los Angeles with more than 137,000 followers whose reporting on the Jan. 6 riot has been cited in Justice Department charging documents. Loder, who uses they/them pronouns, recently had their account suspended, briefly reinstated and then suspended again without a clear reason.
Twitter initially mistakenly marked the account as spam, reinstated it Nov. 23, then suspended it again for ban evasion, according to the company’s notices, which Loder shared in an interview.
Loder said their account was banned soon after being put on a right-wing “target list” of accounts to report for violations of Twitter rules.
On Nov. 25, the account for the antifascist publishing group CrimethInc, which had more than 66,000 followers, was suspended after right-wing activist Andy Ngo tweeted at Musk requesting he ban it. Leigh Young, a member of the group, said that CrimethInc never received an official explanation for the suspension.
“This suggests that the decision to ban our account shortly thereafter was dictated by Musk himself,” Young said.
Twitter did not respond to a request for comment on the suspensions and reinstatements. Requests for comment sent to email addresses associated with Musk bounced back as undeliverable. On Thursday, Musk demonstrated such unilateral decision-making, suspending Ye for “inciting violence” after the rapper tweeted a photo of a swastika inside a Star of David.
The move struck some observers as antithetical to Musk’s previous embrace of free speech.
“Elon ‘absolute free speech’ Musk now draws the line where a number of European countries always have,” tweeted Marietje Schaake, a former member of the European Parliament and the international policy director at Stanford Cyber Policy Center.
The platform’s shift has been enough to drive even some right-leaning people away. Claire Lehmann, the founder of the online magazine Quillette, which emerged in recent years as a popular destination for new libertarian-leaning writers sometimes called the “Intellectual Dark Web,” left Twitter last week. Before deleting her account, shen tweeted that people were being “bullied off” the platform and that followers didn’t equal increased revenue for creators, anyway. “It’s rational to leave,” she wrote.
Celebrities have jumped ship, too. Jim Carrey was the most recent big name to leave Tuesday night, telling his 18.9 million followers, “I love you all so much!” The actor follows a deep bench of famous users who have taken their content off Twitter since Musk’s takeover. Accounts belonging to musician Trent Reznor, acctress Whoopi Goldberg, singer Toni Braxton, musician Moby and TV producer and writer Shonda Rhimes, have all been deleted or sit stagnant. Links that weeks ago pointed to accounts with millions of followers now sit frozen on goodbye tweets, or serve as little grave markers that read “This account does not exist.”
It’s not the first time prominent Twitter users have loudly fled the platform, in many cases, only to reactivate their accounts after a few days or weeks off. And there is some evidence that the high-profile exits are outliers. Data from two independent research firms also found that downloads and activity have grown in the weeks since Musk took over, most notably in the United States.
But data collected by researchers who track online hate speech seems to support the recent claims that the platform has become less a place for “healthy conversation” and more a platform where hate and misinformation spread unchecked. Musk has countered those claims by publishing charts citing Twitter data of volume of total impressions showing hate speech has declined, though he has not detailed how the company reached those conclusions.
He reiterated that claim on Friday, tweeting that the overall reach of hate speech based on the number of times tweets are viewed has declined.
The Network Contagion Research Institute, an independent organization that monitors hostile ideological content online, reported a 500% spike in the use of the N-word on Twitter in the hours after Musk took over. It also pointed out the seeming reward for hate speech on the platform, reporting Ye nearly doubled his follower count after making antisemitic tweets. “Chatter, toxicity, antisemitic conspiracy and new followers escalate with each successive controversy,” the group wrote on Twitter.
Musk has responded to claims that Twitter has become more volatile. Last month, he tweeted that Twitter would follow a “freedom of speech, but not freedom of reach,” policy. “Negative/hate tweets will be max deboosted & demonetized,” he tweeted.
Imran Ahmed, the CEO of the Center for Countering Digital Hate, a nonprofit group, called Musk’s claims “max gaslighting.”
Last week, the group reported that Twitter failed to moderate tweets that leveled racist abuse at athletes competing in the FIFA World Cup. And in an unpublished report provided to NBC News, the group showed tweets promoting hate toward LGBTQ people had been viewed tens of millions of times in the wake of a mass shooting in Colorado Springs, Colorado.
A study from the Center for Strategic Communication at Montclair State University released this week found a dramatic spike in anti-LGBTQ slurs on Twitter following the shooting at the Colorado nightclub, with the terms “grooming” or “groomer” peaking at 885% over their highest level before the shooting.
“We all understand the importance of freedom of speech, but there are dangers to running platforms like Twitter as a ‘free speech absolutist,’” Bond Benton, an associate professor in the School of Communication and Media, said in the report.
According to data from the Center for Countering Digital Hate, one of the most widely viewed anti-LGBTQ tweets came from James Lindsay, a right-wing media personality who takes credit for popularizing the slur “groomer.”
He had previously been banned permanently from Twitter for violating its rules on hateful conduct.