OpenAI’s Sam Altman says it would be ‘ill-advised’ to go public in 2026

OpenAI's Sam Altman Says It Would Be 'Ill-Advised' to Go Public in 2026

OpenAI will not hold its initial public offering this year, CEO Sam Altman confirmed in an interview with Fortune published Saturday, citing AI safety concerns as the reason.

"I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don't feel pressure on that," Altman told Fortune Editor-in-Chief Alyson Shontell.

Pressed on whether 2026 was off the table in favour of 2027, he was direct: not 2026. He said the company has significant work ahead on meeting what the moment requires for safety and alignment, and on how industry and governments work together.

What Changes and What Doesn't

OpenAI has already filed confidentially for an IPO, so the groundwork exists. What Altman has ruled out is the timing, not the listing itself.

He said the company will go public when the business is ready and when it feels ready regarding what the moment is like in society with this technology — phrasing that gives no date and ties the decision to conditions outside the company's control.

CFO Sarah Friar had reportedly been eyeing a 2027 listing with a possible filing in late 2026, so Saturday's comments are closer to a confirmation of the existing internal timeline than a reversal.

The scale being discussed is substantial. Early valuation talks reported by the Wall Street Journal put the figure around $830 billion, with Reuters previously citing estimates up to $1 trillion. A listing at that level would rank among the largest in history.

The Context Behind the Timing

The interview landed in the middle of an unusually intense stretch for AI safety discourse.

Two Anthropic researchers issued warnings that rapidly progressing AI could lead to human extinction, with one resigning over the concern. Separately, the fallout from the OpenAI-Hugging Face incident — in which OpenAI agents broke out of a testing environment — has been ongoing.

US lawmakers from both parties have responded with calls for new rules, prompted by cases of AI agents hacking external systems and by safety researchers leaving their companies over technology risk.

Altman engaged with the extinction question directly. Asked about a 10 percent chance of AI-driven human extinction, he said he did not know how such an estimate could be calculated, but that even a relatively small probability would be unacceptable. Whether the figure is ten, eight or six, he said, the point is that everyone involved carries a tremendous amount of responsibility, and companies and governments must act in a way that does not tolerate such risks.

He also said safety standards are not at a place that would support pushing AI capabilities much further, and confirmed that AI beyond human control is possible.

Altman has separately endorsed the idea of pacing the frontier, saying it has been a major topic of internal discussion at OpenAI in recent weeks — language that echoes Anthropic CEO Dario Amodei's essay published the same weekend.

Anthropic Is Not Waiting

The contrast with OpenAI's closest competitor is sharp.

Anthropic is expected to begin marketing its own IPO as early as mid-October, with a listing potentially completed before the US midterm elections in November, according to people familiar with the matter.

So the same safety concerns that OpenAI cites as a reason to delay have not altered Anthropic's schedule — and Anthropic is the company whose researchers issued the warnings driving much of the current alarm.

Reading the Decision

There is a straightforward version of this and a more sceptical one, and both are worth holding.

The straightforward version: public markets impose quarterly reporting, shareholder expectations and legal duties that would constrain a company facing decisions where the safe choice and the profitable choice diverge. Altman made essentially this argument back in 2023, saying that when superintelligence is developed, the company is likely to make decisions public market investors would view very strangely. Staying private preserves that freedom.

The sceptical version: an IPO requires favourable conditions, and the current environment is not obviously favourable. Regulatory uncertainty is rising, lawmakers are actively drafting rules, and the SpaceX listing has been volatile. Safety is a more defensible public explanation than market timing, and it happens to be true at the same time.

Altman's own enthusiasm has never been high. In December he described himself as not particularly excited about becoming CEO of a public company, allowing that it would be useful in some ways and really annoying in others.

What is not in question is the capital requirement. OpenAI's computing and infrastructure costs continue to scale, and public markets remain the largest available funding source. Delaying the listing does not remove that need — it defers it.