Following his takeover as CEO, the platform has disbanded its trust and safety teams; revoked bans on extremist and dangerous accounts; removed labels informing users that accounts were associated with foreign governments (including Russian and Chinese propaganda outlets); censored journalists critical of Musk; and allowed for what users report, anecdotally, to be a sharp increase in hate speech, online trolling, and harassment. All this marks the decline of Twitter as a trusted platform for news and information.
In search of new sources of revenues, Musk eliminated of Twitter’s key anti-disinformation features: The identification of verified users and authenticated accounts with a blue check mark. This Twitter innovation increased trust on a platform once indispensable for following breaking news. Verification was undertaken by Twitter to identify public figures and public agencies, as well as many journalists and commentators. The check mark was introduced in 2009 to address the problem of impersonator accounts on the platform and denoted that accounts had been verified for authenticity, although the process wasn’t renowned for transparency. In May 2021, the pre-Musk Twitter relaunched its verification program, setting clearer guidelines for who was eligible and who was not. Originally used for public officials, agencies, artists, athletes, and other well-known individuals at risk of impersonation, the blue check mark was rolled out for businesses and brands as well.
Along the way, the blue check mark also became something of a status symbol. “Blue checks” were a kind of Twitter elite, voices worth listening to, a who’s who that included the platform’s most prominent and widely followed users. And if the check mark was so coveted, Musk presumably thought, why not try to monetize it? Users seeking the blue check mark now have to pay for Twitter Blue, but the only authentication that seems to take place is whether the form of payment works. Though Twitter kept the respectable-sounding “verification” label for the new subscription service, any paying user can now obtain a blue check mark. Most of the old blue checks—users whose authenticity was actually verified—had their status taken away if they didn’t pay up.
The rollout of the Twitter Blue subscription service has been chaotic, and a wave of fake accounts with blue check marks immediately emerged. The New York Times reported that in the 24 hours after Twitter changed the rules, at least 11 new accounts with blue check marks began impersonating the Los Angeles Police Department. Someone pretending to be the mayor of New York City promised to create a Department of Traffic and Parking Enforcement and slash police funding by 70 per cent. The Guardian noted that blue-checked accounts appeared posing as the U.S. Internal Revenue Service amid tax-filing season. That Musk calls this service “verification” is an ingenious piece of disinformation itself.
Twitter Blue accounts have fueled the spread of disinformation, especially about the Russia–Ukraine war—a topic on which Musk has repeatedly adopted Kremlin talking points. Fake news amplified by Twitter Blue users (and the Twitter algorithm that increases the visibility of their accounts and individual tweets) include horror stories about alleged organ harvesting in Ukraine and the use of guns destined for Ukraine in recent French protests, according to a report published by BBC Verify on Sunday; the report also traced some of this information back to pro-Kremlin propaganda accounts. The new Twitter Blue system not only amplifies these messages on the platform, but also exploits what many users still perceive as a verification system in order to create an illusion of credibility.
Musk appears to be dismantling the foundations of the platform’s success—and destroying what little keeps social media resilient to information warfare.
Delivering information during breaking global news events used to be Twitter’s bread and butter. Instead, the new Twitter is inimical to democracy. Pay-to-play verification de-amplifies voices unable—or unwilling—to fund a subscription. It has made the already challenging task of identifying authentic accounts and sources of information more difficult.
The program hasn’t exactly been a great commercial success. When blue checks meant that accounts were actually verified by Twitter, their numbers rose from about 5,000 in 2010 to more than 420,000 in 2022. The latter figure represented less than 0.2 per cent of the platform’s almost 240 million daily active users. According to Vox, the problem with Musk’s pay-to-play check mark scheme is that, so far, not enough people are willing to pay. Research suggests that just 4.8 per cent of the legacy verified accounts had signed up for Twitter Blue by April, when Twitter ended the old verification system, and that the majority of Twitter Blue users have fewer than 1,000 followers.
Twitter is now giving away check marks to “important individuals,” but some of the old blue checks don’t want it anymore—presumably to protest the change or avoid being associated with the new pay-to-play status and blue-checked fake accounts. Celebrity journalist Kara Swisher, for example, can’t get rid of hers. Like all of Twitter now, the transparency of the process is subject to Musk’s personal whims. He is also personally funding some of the new blue checks and changing Twitter Blue rules based on size of followership.
The replacement of user verification with Twitter Blue has been described as an “apocalypse,” challenging trust and authenticity on the platform. The removal of free verification of real people and recognized sources of information—and its replacement with a check mark sold to anyone—has increased misinformation, disinformation, and the impersonation of legitimate people and organizations. Musk has added to this by wrongly describing the new pay-to-play system as verification.
If that weren’t bad enough, other changes are similarly devaluing Twitter as a source of information. In the first six months since Musk took over Twitter, Vox reported, “he’s laid off thousands of people, cut Twitter’s valuation in half, released cherry-picked ‘Twitter files’ purportedly showing how pre-Musk Twitter was biased against conservative political views and welcomed formerly banned accounts back into the fold.” A so-called amnesty policy saw 62,000 banned or suspended users reinstated to the platform, many of whom had been banned for violating policies on hate speech, harassment, or abuse. Some 75 of these accounts were megaphones with more than 1 million followers each.
Twitter also made changes to government affiliation rules in April 2023, removing labels from state-controlled or affiliated outlets and ending its visibility filtering system, which previously ensured that these outlets wouldn’t be recommended or amplified. Twitter now algorithmically promotes state-affiliated media, amplifying outlets based in China, Russia, and Iran. Multiple Russian and Chinese state media accounts on Twitter simultaneously began to gain followers after the change. Following months of decline or stagnation in their follower numbers, these propaganda accounts are now thriving on Twitter.
RT Editor in Chief Margarita Simonyan, a well-known face of Kremlin disinformation, tweeted to personally thank Musk “from the heart” for removing her state affiliation label. Simonyan is renowned for her vitriol against the West and is one of the chief propagandists in favor of the Kremlin’s genocidal war against Ukraine. For example, in mid-2022, she remarked that the Russian invasion should obliterate Ukraine, stating that the country “can’t continue to exist. … There will be no Ukraine.”
Twitter users no longer need to actively seek out state-sponsored content to see it on the platform; the platform serves it to them. Many state-affiliated media outlets, particularly in authoritarian countries, exist to exert influence domestically and interfere abroad—including though coordinated disinformation campaigns during democratic elections. Without state media labels, these propaganda outlets with a clear mandate to spread disinformation operate without notification to users that the information is likely biased. No wonder RT’s Simonyan was so grateful.
Musk’s changes are a huge regression from the progress on disinformation and state interference made by Twitter and other social media platforms since the 2014 Russian invasion of Crimea and the 2016 U.S. presidential election. We know that what Musk is doing to Twitter flies in the face of existing research and practice on how to make social media resilient to hate, harassment, and authoritarian governments’ information operations. The new Twitter harms citizens of democracies and benefits autocratic governments.
What makes this even more disconcerting is that Musk has repeatedly adopted Kremlin narratives on Ukraine. Former U.S. National Security Council Russia specialist Fiona Hill wrote that, knowingly or unknowingly, “Elon Musk is transmitting a message for [Russian President Vladimir] Putin.” No stranger to autocratic talking points, Musk has also suggested that China be given partial control of Taiwan. He is not known for making any statements that are critical of the Chinese government.
That Musk calls the new Twitter Blue “verification” is an ingenious piece of disinformation itself.
Musk’s relationship with China is complicated by the fact that his companies (which include Tesla, Starlink, SpaceX, and Twitter) have various relationships, including financial ones, with the Chinese government. He has repeatedly heaped praise on China and the Chinese Communist Party. In his most recent visit to China in May and June 2023, Musk was uncharacteristically quiet on social media but visited Tesla’s factory in Shanghai and found time to compliment China’s technological development, continuing a long pattern of praise for China. He also described the economies of the United States and China as “conjoined twins” and stated his opposition to any moves that might undo that.
Of course, China is a vital part of Tesla’s electric vehicle empire—as a market, a production site, and a source of financing for the company. The Shanghai plant reportedly produced more than 700,000 Model Y and Model 3 vehicles last year, more than half Tesla’s global output. Musk’s visit follows those of several other U.S. CEOs, including Apple’s Tim Cook and General Motors’s Mary Barra, but Musk is in a unique position. He is financially dependent on China and simultaneously making decisions that smooth the way for Beijing’s information operations.
There are even more signs that the platform is not fighting disinformation and state-sponsored interference. In May 2023, Twitter pulled out of the European Union’s voluntary code to fight disinformation. Twitter hasn’t published a transparency report since Musk took ownership. Self-reported data also shows that Twitter has complied with hundreds of government orders for censorship or surveillance of users in Turkey, India, and other countries.
The dissolution of Twitter’s Trust and Safety Council and harassment of former trust and safety chief Yoel Roth further highlight damage to the platform’s integrity and safety. Last month, Twitter’s new trust and safety chief also resigned, the second to do so since Musk took over. The slashing of Twitter’s global workforce from 8,000 to 1,500, including many working in the trust and safety teams, has resulted in an increase in complaints and lawsuits. Twitter has allegedly failed to pay more than 1 million Australian dollars (more than $650,000 in the United States) for office space and management in London, Dublin, Sydney, and Singapore.
Users are increasingly subject to online harassment and hate on Twitter. A nationally representative survey of online hate and harassment experienced by U.S. users in the 12 months ending with April 2023 found that the share of users who said they experienced online harassment on Twitter rose to 27 percent, compared to 21 percent one year before. The Center for Countering Digital Hate, a nongovernmental organization based in Washington and London, found that Twitter failed to act on 99 percent of tweets by Twitter Blue subscribers reported to Twitter the using the platform’s own tools for flagging hateful conduct.
In Australia, rising complaints to the Australian eSafety office and reports of the toxic content remaining on the platform suggest that Twitter is probably not enforcing its own rules. Last month, Twitter defended its handling of disinformation in Australia, but at the same time reduced reporting on content removal and user-data disclosures. Australia’s eSafety commissioner issued legal warnings to Twitter last month, saying, “Twitter appears to have dropped the ball on tackling hate.” This matches reporting from Twitter insiders who said the platform was no longer equipped to protect users from trolling, state-coordinated disinformation, and child sexual exploitation. Insiders report that all that remain are automated detection systems. It’s not possible to confirm this with Twitter directly, since media inquiries receive an auto-reply consisting of a poo emoji.
Twitter has long been a source for journalists and media outlets, as well as a source of news and current affairs for most users. However, Musk appears to be dismantling the foundations of the platform’s success—and destroying what little keeps social media resilient to information warfare. We already know that Twitter’s algorithms amplify certain political voices over others. Twitter’s recent elimination of verification, abolishment of the government affiliation label, reinstatement of previously banned accounts, significant reductions in trust and safety staff, and decreased focus on fighting disinformation have created an algorithmically fueled propaganda megaphone.