Twitter’s underlying software is reportedly showing signs of showing signs of cracking for some users, without enough engineers to maintain it. Cybersecurity and privacy experts are worried that it’s becoming a “Wild West” when it comes to data breaches and vulnerability to hacking.

So far there hasn’t been a mass exodus — users in politics and media are watching and waiting, and Twitter has remained largely reliable as a real-time news platform. But Schmidt says she and her colleagues are trying to determine whether it’s safe for their clients to keep advertising, or even stay on Twitter at all.

Several Washington communications veterans contacted by POLITICO said versions of the same thing: Twitter now needs to be treated with great caution by anyone concerned with their public image.

“Whether it’s a politician, or candidate, or a business, they need a platform that has credibility and that is stable and aligns with their values,” said Sean Higgins, a veteran of the D.C. political communications scene and associate vice president at Precision Strategies. “So far, Twitter hasn’t demonstrated an ability to provide any of those things under Elon Musk’s leadership — and that’s a problem.”

Central to the anxiety is that in just two weeks since Musk’s takeover, Twitter has largely dismantled its previous system for verifying accounts, launching a short-lived subscription service that allowed “verified” status to anyone willing to pay $8 and generating a mass of fake accounts for government agency heads, companies and politicians.

It launched, canceled and re-launched a new “official” badge to designate real accounts, but that appears to be unevenly applied. So does the “United States Government organization” tag, which appears on the Twitter account for the Department of Defense, but not the White House or other crucial agencies like FEMA.

As Twitter works to fix its content moderation problems with a skeleton crew, Higgins warned that businesses and advertisers did not have “very much patience.”

Musk has argued that all of this attention has been good for Twitter, and says it is more says it is more popular among users than ever. But what looks to some like a satirical free-for-all looks to others like the rapid unraveling of a platform that had, over the past several years, become a reliably policed, if still occasionally toxic, public forum.

At the center of Twitter’s rapid operational changes is Musk’s stated goal to “disrupt” mainstream media’s “oligopoly on information” by “elevating citizen journalism” — and, importantly, making $8 per paid subscriber along the way. But before that happens, the platform owner needs to contend with the steep real-world consequences of breaking the machine that it had built to ensure reliability.

Chris Riotta, a cybersecurity writer for D.C. federal technology trade publication FCW put it this way: “Elon Musk’s decision to monetize verification on Twitter … marks the end of an era for social media, where Twitter users could easily confirm whether a post is trustworthy.”

Insulin manufacturer Eli Lilly’s stock value dropped sharply after one fake tweet that its insulin was now free. And Sen. Ed Markey (D-Mass.) sent Musk a strongly worded letter demanding an explanation after a Washington Post reporter posed as the politician just to show how easy it was.

Musk characteristically took Markey’s critique as an opportunity to write a trolling response, tweeting: “Perhaps it is because your real account sounds like a parody?”

Markey did not take the response kindly. “Fix your companies. Or Congress will,” tweeted the Senator, warning Musk that one of his companies is under an FTC consent decree.

The rising pile of incidents point to a rollback of years of efforts to turn Twitter from a frivolous chatroom for the tech world insiders to the most important and reliable real-time online news feed.

A D.C. crisis communications professional with four years of experience supporting clients in the tech industry said, “Between the contentious brand Musk has created for himself on social media and the public tendency to use humor to cope with current events, people may be beginning to lose hope that the issue is solvable.”

D.C. stalwarts like Tom Wheeler, a Brookings visting fellow and former FCC chair, have already mentioned the heightened regulatory risk facing both Musk and Twitter. Wheeler added that the surprise awaiting the new platform owner is how legislators and agencies who rely on Twitter for their messaging will react to “the potential that such capricious actions could impact their political brand.”

Mark MacCarthy, a Brookings senior fellow, had a blunt estimation of Twitter’s new plan to sell verification labels. He called it “stupid.” MacCarthy is an adjunct Georgetown professor of communication and a former staff member of the House Energy and Commerce Committee. Skeptical that a “market” in verification could ever work, MacCarthy said, “Musk has to go back to the hard work of weeding out fake accounts using signals, judgment, context and intuition. Which takes some of the very people he has driven from the company.”

Despite reassurances from Musk himself earlier this week, digital advertisers seem to concur with MacCarthy. Many have pulled back on Twitter ad spending as they wait out the confusion.