This Tuesday marked one whole year since Twitter accepted Elon Musk’s $44 billion offer to buy the social network. Many memorable events marked the 12 months that followed—but one pattern, among all, stands out: the sheer regularity with which Musk says one thing and proceeds to do the complete opposite.

This became most clear recently as the debacle over blue check verification came to a head and the company reversed course on labelling media organizations that receive public funding.

“Throughout his career, Musk has had an almost pathological need to promise grand visions and make himself the centre of attention,” says Jesse Lehrich, co-founder of Accountable Tech, an advocacy group. “He’s very Trumpian in his need to capture media attention with constantly-shifting promises, which everyone in the media covers, and then it never happens.”

Twitter’s content moderation

Shortly after completing his acquisition of Twitter in October 2022, Musk said he would establish a “content moderation council” to advise him on controversial decisions, such as whether to uphold former President Donald Trump’s suspension from the platform. (Musk said he would not reinstate the former president without consulting this council.)

But the council was never established and in November Musk unilaterally revoked Trump’s suspension from the platform. He fired many members of Twitter’s platform safety team just days before the U.S. midterm elections.

Under Musk, Twitter has also removed bans on dozens of accounts including Neo-Nazis and disbanded the platform’s already-existing Trust and Safety Council, which was populated by civil society groups.

Suspension of ElonJet

On Nov. 7, 2022, Musk tweeted that he would not ban an account that monitored the movements of his Gulfstream G700 private jet in real-time.

The account would often post tweets revealing short flights—the 40 miles from San Jose to San Francisco, for example—that hurt the electric vehicle billionaire’s environmentally-friendly image.

“My commitment to free speech extends even to not banning the account following my plane, even though that is a direct personal safety risk,” Musk tweeted on Nov. 7. In December, Musk banned the account, claiming it had violated Twitter’s doxxing policy.

Support for free speech

Before his acquisition of Twitter, Musk said he would like to allow all legal speech to remain on the platform. “I hope that even my worst critics remain on Twitter because that is what free speech means,” he tweeted in April.

By December, he had suspended at least nine prominent journalists from the platform, ostensibly for retweeting or publicizing ElonJet’s tweets about the location of his private jet. The journalists’ accounts were later reinstated.

At around the same time, the Twitter account of rival social network Mastodon was suspended from the platform, and links to Mastodon servers were prevented from being posted on Twitter. The platform later reversed course on those decisions.

Stepping down as CEO

In late December 2022, Musk ran a Twitter poll asking whether he should step down as the head of Twitter. “I will abide by the results of this poll,” he wrote. More than 17.5 million votes were cast, and “Yes” won with a tally of 57.5%. More than four months on, Musk is yet to step down.

“I will resign as CEO as soon as I find someone foolish enough to take the job!” he tweeted on Dec. 21.

Defeating spam

Musk has overhauled Twitter’s verification program, whose blue checkmark was formerly a signal that a public figure’s account was owned by who it was claiming to be.

Musk has replaced that invite-only system with Twitter Blue, an $8 per month subscription accessible to anybody. The billionaire said the change would be “essential to defeat spam.

” But soon, some users began paying for verification in order to impersonate the accounts of famous brands. A fake (but verified) account pretending to be the pharmaceutical company Eli Lilly tweeted “We are excited to announce insulin is free now,” forcing the company to proclaim that this was not true.

Brands who had spent big on advertising on Twitter began deserting the platform, concerned about the risk of impersonation.

U-turning on blue checks

Musk’s vision for Twitter Blue has taken many twists and turns. He announced in March that only accounts that subscribe to Twitter Blue would appear in Twitter’s main timeline, known as the For You page. “[This] is the only realistic way to address advanced AI bot swarms taking over,” Musk wrote.

Days later he backtracked, saying he “forgot to mention” non-paying users’ tweets would appear as well. On April 20, the blue checkmarks of “legacy” verified accounts predating Musk’s tenure at Twitter disappeared, but then began reappearing on the accounts of several celebrities who had publicly stated they would never pay for the service, including LeBron James and William Shatner.

Musk then tweeted that he was paying for several celebrities’ blue checkmarks personally, prompting some to observe that he could be exposed to legal risk for suggesting that celebrities were endorsing a paid service when in fact they were not.

Backpedalling on media labelling

Musk received pushback from media organizations including NPR, PBS and the BBC after affixing “state-funded media” labels to their accounts without warning.

The organizations said that the labels were misleading, as they suggested a lack of editorial independence from the government and implied full reliance on government funds despite their commercial activities.

NPR said it was leaving the platform due to the labelling. Musk later U-turned, removing the labels. In the process, he also removed similar labels affixed to Russian state propaganda outfit RT, and China Daily, which is affiliated with the Chinese Communist Party.

Around the same time, it appears, Twitter also removed longstanding limits on the virality of posts from those accounts, leading to spikes in engagement, according to the Atlantic Council’s Digital Forensic Research Lab.

Concern for AI safety

In March, Elon Musk signed an open letter warning of the severe risks of companies racing toward building ever-more-powerful AIs. The letter called for a six-month pause on training new AIs above a certain power threshold, and argued that companies were in an “out-of-control race” to beat one another was leading to dangerous corner-cutting.