Dario Amodei, co-founder and CEO of Anthropic, speaks to “The Circuit with Emily Chang” at Anthropic’s headquarters in San Francisco, California, on Thursday, April 30, 2026.
Jason Henry | Bloomberg | Getty Images
Anthropic’s path to an IPO just got tougher.
While Claude’s creator meets with prospective investors ahead of what will likely be a historic debut, co-founder and CEO Dario Amodei is pushing a concept that seems at odds with these ambitious efforts: slowing down.
Anthropic, valued at $965 billion earlier this year, secretly filed an IPO prospectus in June and was widely expected to list its shares as early as next month. Meanwhile, concerns about the power of advanced AI models have grown in recent weeks, spilling over into the mainstream as more researchers warn of the potential threat of human extinction.
Against this backdrop, Amodei wrote an essay over the weekend urging the AI industry to slow the pace of model development and proposing a three-step plan to slow the rate at which models improve without “sacrificing commercial advantage or America’s lead in AI.”
It’s the latest challenge facing public market investors trying to figure out how much they’re willing to pay for part of a five-year-old company that’s already one of the world’s most valuable and could seek a $2 trillion valuation in an IPO. While Anthropic may have to accept a hit to its revenue growth, some experts say the intentional slowdown could help Anthropic position itself as a responsible actor, avoid future liability and address the public backlash against AI occurring across the United States.
“I don’t know if investors will necessarily see that as a negative,” Gil Luria, an equity analyst at DA Davidson, said in an interview. “Unless a company really says, ‘Okay, we’re not going to do an IPO, we’re not going to use any more compute, we’re not going to train any more models.'” That’s not what they’re saying. ”
After Business Insider first reported the choice, CNBC confirmed that Anthropic had chosen Nasdaq as the exchange for its potential IPO.
Amodei on Saturday suggested model companies open their doors to third-party evaluators, frontier companies to establish “common safety standards” and democracies to work with authoritarian governments “to the extent possible.”
His essay comes after several industry researchers issued stark warnings last week about the technology’s increasing potential to cause catastrophic harm.
OpenAI CEO Sam Altman expressed support for Amodei’s proposal, as did OpenAI CEO Elon Musk. space xowns Grok Creator xAI. SpaceX went public in June in the largest IPO in history and is now valued at $2 trillion. OpenAI secretly filed its IPO prospectus, but has come under fire in recent months after its models escaped containment, accessed the open internet, and infiltrated open source developer platform Hugging Face.
“This is not a wise time to go public,” Altman said in an interview with Fortune magazine, reiterating that OpenAI does not intend to pursue an IPO until next year. Finance Director Sarah Friar told employees at an all-hands meeting last month that the AI Institute “will become a publicly traded company in 2027.”
Lise Bayer, a partner at IPO advisory firm Class V Group, said she doesn’t think recent concerns that “we could wipe you out” will affect the timing of an IPO, but said valuations could change.
“The bet here is long-term, and we are now softening our thinking about managing the technology,” Bayer said in an email. “The dramatic growth and potential of these companies is now more public, coupled with potentially very serious concerns and risks that will likely persist whether the IPO takes place in the fourth quarter, next year, or whenever.”
Anthropic and OpenAI declined to comment for this article.
“I don’t know why growth is slowing down.”
As CNBC previously reported, Anthropic’s annual sales reached $65 billion in July, about seven times more than a year ago. The Financial Times reported on Sunday, citing people familiar with the matter, that Anthropic has told some shareholders that it expects to post an operating profit for the second straight quarter this fiscal year.
Matt Murphy, a partner at Menlo Ventures and an Anthropic investor, called the growth rate “outrageous” and said going public would force Anthropic to be more transparent about its business and could help improve the gloomy public opinion of its technology.
“I don’t see any reason why growth would slow or any other reason to wait,” Murphy told CNBC.
More than half of Americans say they are more concerned than excited about the growing use of AI in daily life, up from 37% in 2021, according to a recent report from the Pew Research Center. And trust in AI executives is getting worse, according to a CNBC Generation Lab survey of 18- to 34-year-olds. More than 75% of respondents said they did not trust Mr. Amodei to act responsibly, and about 70% expressed that opinion about Mr. Altman.
“The accountability that comes with being a public company is a big concern for a lot of people, so some would argue that the sooner the better,” said Bayer of Class V Group.
Brad Gerstner, CEO of Altimeter Capital, which is also an investor in Anthropic and OpenAI, said in a post on X on Saturday that it is “critically important” to bring more “transparency, oversight, accountability” and participation to AI companies. He said Anthropic is likely to move forward with an IPO.
“The market knows how to price risk. See SpaceX,” Gerstner wrote. “The appetite to invest in AI leaders is very high.”
Gerstner’s post came a day after he criticized industry researchers’ public statements as “hyperbolic scare tactics” that are “hidden behind political objectives” in an interview with CNBC.
Many are skeptical about Amodei’s latest positioning. One argument is that Anthropic benefits from more stringent standards because it currently has cutting-edge models and makes money selling services like Claude Code that leverage those models.
“It could actually favor Anthropic and OpenAI if smaller competitors cannot afford the rigorous safety, reputation, and security investments required for frontier-level models,” Gartner analyst Arun Chandrasekaran told CNBC in an email.
DA Davidson’s Luria agreed, saying he believes Anthropic and OpenAI are engaging in “monopolistic behavior.” Wired reports that OpenAI has asked members of Congress for guidance on whether a coordinated industry-wide slowdown violates antitrust laws.
“I’m very curious about what Anthropic and OpenAI are doing,” Luria said. “It feels more and more like pulling a ladder.”
What about the rest of the technology?
There are other reasons for technology investors to be concerned about the pace of development at OpenAI and Anthropic. That’s because these companies are shouldering huge AI infrastructure spending.
Anthropic has signed a flurry of multibillion-dollar computing deals this year, including a deal with Nscale. advanced micro device, space xand google. OpenAI told investors in February that it was targeting total computing spending of roughly $600 billion by 2030. Both companies Nvidia’s Graphics processing unit.
“We want to understand how the combination of frontier training, post-training, and inference changes as safety management becomes more integrated,” said Ro Tony, managing partner at Plexo Capital and Anthropic investor.
PitchBook analyst Harrison Rolfes is more concerned about slower growth. He said valuations for model companies should now be discounted, mainly because it is difficult for investors to believe that their technology can be safely commercialized.
“Is the first thing you want to do as a publicly traded company to address a lot of security issues and vulnerability issues?” Rolfes said. “No, we probably want to focus on expanding into all the markets we promised all our investors.”
Gene Munster, managing partner at Deepwater Asset Management, told CNBC that any perceived slowdown of any kind would be negative because the market is “undertaking an exponential continual improvement of the model.”
Still, Munster predicted that “nothing will change and the AI leapfrog game will continue.”
“The long-term opportunity for AI is too great to slow down,” Munster said. “I think the comments were aimed at relieving regulatory pressure.”
WATCH: Anthropic looks set to beat OpenAI to IPO, says FirstMark’s Rick Heitzmann
