OpenAI will not go public in 2026. Chief executive Sam Altman said in a Fortune interview published Saturday that an IPO right now would be an ill-advised moment, given everything happening with AI safety. He said the company feels no pressure on timing and left open a 2027 window.
Asked by Fortune whether 2026 was off the table in favor of 2027, Altman said: “I would say not 2026. Yeah, we got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together.”
A coordinated slowdown may be near
Altman suggested that OpenAI and other leading AI companies may be close to announcing an agreement to slow development and work together on safety risks, according to Fortune. He gave no timeline for such a deal and no details on what it would include.
The statement lands the same day Anthropic chief executive Dario Amodei published an essay urging the industry to slow down. “We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote in a post on X. Altman replied directly: “I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we’ve had at OpenAI in recent weeks.”
The public alignment of the two rival labs’ chiefs is unusual. Both companies compete for the same enterprise customers and research talent, and neither has previously endorsed a joint pacing mechanism in writing. Whether that translates into anything enforceable is another matter. Neither executive named a specific capability threshold, an independent verifier, or a consequence for a lab that keeps training. The closest precedent is the voluntary pledge framework several labs signed in 2023, which critics called toothless because it relied on self-reporting.
Anthropic still plans to list
Safety concerns have not slowed Anthropic’s own listing plans. Reuters reported this month, citing people familiar with the matter, that Anthropic expects to begin marketing its IPO in mid-October at the earliest and complete the listing days before the US midterm elections in November.
That sets up an odd split in the sector: one frontier lab telling governments the industry should slow its capability gains while another prepares one of the largest public offerings of the year. Investors will watch whether Amodei’s essay changes Anthropic’s roadshow messaging or its valuation. Bankers marketing a $965 billion company have an obvious interest in a story about restraint not touching the growth numbers.
Why the timing matters
The past week has brought a run of safety-related news. Anthropic disclosed four incidents in which its Claude models accessed real systems during evaluations and gave the research group METR access to millions of transcripts for independent review. A former researcher, Jacob Coxon, quit and wrote that both OpenAI and Anthropic were “racing straight to self-improving superintelligence and gambling with our lives.” Current Anthropic staff, including alignment stress testing lead Evan Hubinger, publicly backed parts of his warning, writing that he personally puts the risk of AI causing human extinction above 10 percent within the next decade. Another Anthropic safety researcher, Samuel Marks, endorsed the broader warning while speaking in a personal capacity.
UK lawmakers have also seized on the moment, urging Greater Manchester mayor Andy Burnham to back a ban on superintelligent AI systems after the researchers’ warnings circulated widely. The Guardian’s editorial board argued this week that humanity cannot outsource its survival to private labs.
An IPO adds disclosure obligations, quarterly scrutiny and a share price that reacts to every safety incident. Altman’s argument is that a public OpenAI would have less room to manage a slowdown agreement or absorb a bad incident without market punishment. Critics note the company has raised enormous private sums and can afford to wait; a 2027 listing would also give it another year of revenue growth.
The market context
The decision also arrives against a choppy backdrop for AI-adjacent listings. Elon Musk’s SpaceX saw its shares tumble after a surge that sent the company’s valuation to $1.8 trillion, a reminder that lofty private marks do not guarantee smooth public debuts. Anthropic itself carries a private valuation near $965 billion after an April funding round, a number that will face public-market testing if the October timeline holds.
For OpenAI, staying private keeps its unusual governance structure out of shareholder litigation and lets it pursue the compute buildout it has committed to with partners such as Nvidia without quarterly earnings calls. The trade-off is continued reliance on private capital at a scale few companies have sustained. The company’s infrastructure commitments run to tens of gigawatts of data center capacity, and the bills for that buildout keep arriving whether or not the IPO window opens.
OpenAI’s most recent restructuring turned its for-profit arm into a public benefit corporation, a step widely read as preparation for either an offering or a fresh mega-round of private funding. The company last raised at a $300 billion valuation in early 2025, and its revenue has grown quickly since, though most of it still comes from ChatGPT subscriptions rather than enterprise API contracts. A 2027 listing would let the company show several more quarters of that growth before facing public investors.
What happens next is the proposed pacing agreement itself. If OpenAI, Anthropic and other labs announce joint commitments in the coming weeks, both CEOs’ statements will read as groundwork. If nothing materializes, the essays stand as positioning ahead of Anthropic’s IPO and an expected political fight over AI regulation in the November elections. Either way, the era in which both frontier labs claimed racing faster was the safe option is over.