The Washington PostDemocracy Dies in Darkness

Elon Musk’s first big Twitter product paused after fake accounts spread

Twitter accounts impersonating celebrities and politicians spread on the site after the company rolled out paid check marks

Twitter accounts impersonating celebrities and politicians spread on the site after Elon Musk rolled out a paid subscription option that adds a blue check mark on an account. (Patrick T. Fallon/Bloomberg News)
9 min

Twitter on Thursday night pressed pause on Elon Musk’s first major product launch — a paid-for blue check mark — after misinformation flooded the site.

Twitter rolled out Twitter Blue on Wednesday, letting users who provided an Apple ID and a phone number pay $7.99 a month to attain verification. Previously, that badge was only available to people when the company had verified their identity, often public figures and brands.

Almost immediately, users started taking advantage of the new tool. Accounts were created impersonating politicians including President Biden and celebrities, as well as other notable people. Several also surfaced purporting to be brands, announcing fake news.

Twitter temporarily disabled sign-ups for the new service Thursday night, according to an internal note viewed by The Washington Post, to “help address impersonation issues.”

But in several cases, the damage was already done, and some fake accounts were still active as of Friday.

Twitter’s content moderation chief quits

On Friday afternoon, Sen. Edward J. Markey (D-Mass.) sent a letter to Musk asking several questions about the blue check mark subscription program. A Washington Post columnist set up an account impersonating Markey this week, with the senator’s permission, and paid for a blue check mark.

“Apparently, due to Twitter’s lax verification practices and apparent need for cash, anyone could pay $8.00 and impersonate someone on your platform,” Markey wrote. “Selling the truth is dangerous and unacceptable.”

Twitter did not respond to requests for comment.

On Wednesday, a fake account purporting to be basketball star LeBron James falsely tweeted that the athlete was requesting a trade. Another fake account with a blue check mark pretending to be former president George W. Bush tweeted “I miss killing Iraqis.”

And a post from a fake account pretending to be pharmaceutical giant Eli Lilly was shared 3,000 times and remained online for several hours Thursday afternoon. The tweet read: “We are excited to announce insulin is free now.” An Eli Lilly spokesperson told The Post on Thursday they “are in communication with Twitter to address the issue.”

The decision to pause a signature new product under Musk — the world’s richest man who is also a Twitter super user — marks two weeks of chaos under the new ownership. Musk, who already counts himself as CEO of companies including Tesla and SpaceX, has moved quickly to implement changes and has had to backtrack multiple times in recent days.

Twitter’s paid verification service is here. What you need to know.

Last week, Musk laid off roughly half of Twitter’s 7,500 staff members, raising concerns about the company’s ability to police misinformation and other harmful content on the site. Over the weekend, the company tried to hire some of them back.

Civil rights groups called on advertisers to suspend their campaigns on Twitter, and many have. Additionally, a string of executives have left the company, including — and perhaps most notably — the company’s head of content moderation, who participated in a Twitter Spaces public meeting with Musk and advertisers on Wednesday.

Advertisers fleeing, workers in fear: Welcome to Elon Musk’s Twitter

Musk also ordered staffers to return to the office, reversing a policy at the tech company that all workers could remain remote — and making more departures likely.

During a meeting with employees held shortly after the return-to-office policy was announced, a worker asked Musk if the company had a specific focus for the next few months.

“We just definitely need to bring in more cash than we spend,” Musk answered, according to a transcript of the meeting published by the Verge. “If we don’t do that and there’s a massive negative cash flow, then bankruptcy is not out of the question. That is a priority. We can’t scale to 1 billion users and take massive losses along the way.”

Twitter Blue is Musk’s first major product change. The initial rollout of the overhauled verification system was dialed back as Musk expressed concern over its design.

Inside Twitter, workers scrambled to address the problems unleashed by the launch.

Twitter has a risk evaluation team that vets new product launches for anticipated issues in compliance and compatibility. The team that built Musk’s version of Twitter Blue did not follow its internal risk evaluation process, however, because of the expedited timeline, and it didn’t make key upgrades equipping the content moderation side to flag problems, according to three people familiar with the process, who spoke on the condition of anonymity to candidly describe the matter.

The risk evaluation team was laid off, one of the people said, leading to a frenzy of last-minute defenses. Following the launch, there were major bugs with tools that slowed down or prevented moderation.

A key problem, the people said, was Twitter internal staff’s inability to immediately distinguish newly awarded check marks from those of official accounts that had been given them under the legacy verification system. That stemmed from a lack of compatibility between the new Twitter Blue system and Twitter’s existing moderation tools, the people said.

Press Enter to skip to end of carousel
Elon Musk and Twitter
‘The bird is freed’
Elon Musk completed the acquisition of Twitter for $44 billion on Oct. 27. A lot has happened since then as Musk moves to overhaul the company, taking it private and firing top executives and handing half of Twitter’s workers pink slips.
Before the billionaire sealed the deal, an analysis of his 19,000 tweets showed his complicated relationship with Twitter.
Content moderation
Musk stated his desire to own the social media company so that there’s “an inclusive arena for free speech.” But not long after Musk took control, the Twitter trolls went on a rampage.
In an attempt to reassure advertisers, Musk said during a Twitter Spaces he had noted concerns about hate speech and misinformation on the platform. However, concerns continued to grow after top executives, including those leading content moderation efforts, quit.
The new blue check marks
Paid verification was the first product to launch under Musk’s leadership, and its roll-out was anything but smooth. The $7.99 subscription allowed any user to have a sought-after blue check mark, but after an explosion of impersonation accounts the program was paused. The backlash to the flood of misinformation caused some brands to pull their advertising from the platform.
Check your Twitter privacy settings if you’re worried about your data.
Be hardcore
Musk issued the remaining staff an ultimatum: commit to a new “hardcore” Twitter with longer hours and no remote work or leave the company with severance pay. Hundreds are said to have made the decision to leave the company.
The number of likely departures prompted Musk to ease his return-to-office edict and managers to meet to decide which engineers to ask back, causing many to question whether Twitter is on the verge of shutting down.


End of carousel

These types of fast rollouts of products were particularly concerning to privacy staffers, some of whom quit Thursday. They said they needed full security reviews required under an order Twitter entered with the Federal Trade Commission earlier this year, following allegations that the company deceptively used phone numbers and other personal information for advertising purposes.

Regulators in Europe, which has more stringent data protection rules than the United States, were closely monitoring the rapid developments at Twitter on Friday. The Irish Data Protection Commission has scheduled a meeting with Twitter officials on Monday, DPC spokesman Graham Doyle told The Post.

The Irish regulators plan to speak with Twitter about the recent departure of the company’s data protection officer, Damien Kieran. European Union laws require companies to have a designated officer on staff, and the Irish regulators were not notified of the data protection officer’s departure before it was reported in the media, Doyle said. The Irish regulator also plans to press Twitter officials about whether key decisions about the processing of E.U. users’ personal data are still being made in Ireland after the departures and layoffs.

Currently Ireland acts as the lead data protection authority overseeing Twitter in Europe. But if the company is no longer making key decisions there — a designation known as “main establishment” — Twitter may have to answer to individual regulators in all 27 E.U. member states, which would introduce a greater compliance burden for the company. The United Kingdom’s Information Commissioner Office also issued a statement Friday, saying that it was continuing to “monitor the situation with Twitter.” It encouraged “anyone with concerns” to report them to the office.

Overnight, Musk tweeted that the site hit an all-time high of active users on Thursday.

Musk took issue with account impersonations last weekend when some changed their name on the site and pretended to be the billionaire. By Thursday, he had tweeted a link to updated Twitter rules and said that “accounts engaged in parody must include ‘parody’ in their name, not just in bio.”

While Twitter Blue is being paused, existing users will still have access to the subscription features, the internal Twitter note said.

In one example of abuse, an account with a blue check mark pretending to belong to Arizona gubernatorial candidate Kari Lake tweeted “I have WON. I decided it so it is truth,” on Thursday.

But the account, with the handle @TheRealKariLake, is not the candidate’s official account. And the race for Arizona’s next governor remains too early to be called — Lake, the Republican candidate, is locked in a close race with Democrat Katie Hobbs.

Users can click on a blue check mark and learn whether an account paid to be verified or was part of Twitter’s legacy program, but it’s otherwise difficult to distinguish. (The Post also found that there appeared to be a bug in the pop-ups that describe the blue check marks — sometimes showing accounts as “notable” when they were instead paid.)

There appear to be other bugs with the new service — the fake Lake account was showing up with a blue check mark for some users, but not others.

We got Twitter ‘verified’ in minutes posing as a comedian and a senator

The fake accounts for James, Bush and Eli Lilly were suspended, but some impersonation accounts with blue check marks for other companies and prominent figures remained online as of Friday morning.

In addition, the company said in its internal note that it would add a gray “Official” label to advertisers’ accounts.

Earlier this week, the company appeared to be rolling out that second label to indicate if accounts are official, but quickly rolled it back.

Musk tweeted on Wednesday he had “killed it,” and a Twitter executive clarified later that the company was focusing on using the badges for “government and commercial entities” instead of individuals.

“Apart from it being an aesthetic nightmare when looking at the Twitter feed it is simply another way of creating a two-class system,” Musk said during the Twitter Spaces on Wednesday. “It wasn’t addressing the core problem of there are too many entities that would be considered official or have legacy blue check marks.”

The account impersonations could cause further rifts between brands who advertise on Twitter and the company. Some stock prices of affected companies sank on Friday, though it was unclear the drops were tied to the fake tweets.

Even real, official accounts took note of the chaos Friday on Twitter. The official account for Washington state’s Department of Natural Resources tweeted, “Update: The Twitter wildfire is at 44 billion acres and 0% contained.”

Faiz Siddiqui and Drew Harwell contributed to this report.