Sam Altman: no OpenAI IPO in 2026, calls timing 'ill-advised'

Altman's comments came in a wide-ranging, 45-minute interview with Fortune that touched on the Hugging Face hacking incident, recursive self-improvement, and the possibility of building an AI beyond human control. On that last point he said it was 'absolutely' possible, but pledged to take steps to prevent it, even if that meant pausing training, and added that 'there are risks we should not be able to incur on behalf of humanity.'

The IPO remark fits that same posture: Altman tied the decision to safety rather than to markets or internal readiness, saying OpenAI has 'a lot of stuff to do' and does not feel pressure. He also said the company had 'said for a long time' it would go public when ready — so the news is less a reversal than a firm date removed from the calendar.

What the timing ultimately hinges on, based on Altman's own framing, appears to be how the safety situation develops rather than any financing need, since he said OpenAI feels no pressure to list. For employees and investors waiting on liquidity, that leaves the eventual window open but unspecified.