Sam Altman rules out 2026 OpenAI IPO and says AI beyond human control is possible

No OpenAI IPO in 2026

OpenAI will not go public next year, and its chief executive believes listing the company right now would be a mistake. In an interview with Fortune, reported by The Verge, Sam Altman ruled out a 2026 IPO and tied the decision directly to the safety questions hanging over the AI industry.

"We're not rushing into an IPO. I actually think that, given everything happening with safety, this would be, right now would be an ill-advised moment to go public," Altman said, according to The Verge. He added that OpenAI feels no pressure on the timing, that the company has long said it will list when it is ready, and that the answer for now is simple: "not 2026," because "we've got a lot of stuff to do."

AI beyond human control is a real possibility

The same conversation turned to more existential territory. According to The Verge's account of the roughly 45-minute interview, Altman was asked whether it is possible to build an AI beyond human control, and he answered "absolutely" — while pledging to act to prevent it, up to and including pausing training.

"There are risks we should not be able to incur on behalf of humanity," Altman said, as quoted by The Verge.

The framing matters. This is not a safety researcher speculating about worst cases; it is the CEO of one of the world's leading AI companies stating plainly that losing control is a genuine possibility rather than science fiction. His answer pairs that admission with a commitment to a costly safeguard — halting training — rather than a dismissal of the concern.

Hugging Face hack and recursive self-improvement also came up

The Verge notes that the Fortune interview also covered the recent Hugging Face hacking incident and recursive self-improvement, the idea of AI systems improving themselves. The outlet's dispatch was brief and did not detail what Altman said on either subject, so the two headline takeaways from the conversation remain the IPO timeline and his comments on controllability.

Why it matters

Altman connected the two threads himself, at least implicitly: the "everything happening with safety" he cited as a reason to delay going public is the same landscape in which he says uncontrollable AI is a real possibility.

On the market side, the statement draws a firm boundary around near-term speculation about an OpenAI listing, and it offers an unusual rationale. Companies typically delay IPOs because of market conditions; here the CEO is delaying, in his own telling, because of the state of the technology and the work still left to do.

On the policy side, a leading lab CEO saying on the record that AI beyond human control is "absolutely" possible — and that he would stop training to prevent it — feeds directly into arguments about how frontier AI should be governed. Safety advocates can point to it as confirmation that the risks are taken seriously at the top; critics can note that a voluntary willingness to pause is not an enforceable commitment, and that outsiders have no way to verify whether such a pause would ever actually happen. Either way, the words came from the person running one of the most influential AI companies in the world.