Altman Rules Out OpenAI IPO in 2026, Calls Going Public 'Ill-Advised'

OpenAI will not go public in 2026, CEO Sam Altman said, calling the idea "ill-advised" for a company still working through the early stages of artificial general intelligence. Altman made the remark in a 45-minute interview with Fortune, where he fielded questions on a wide range of topics, from a recent security breach at Hugging Face to the long-term risks of building systems that outstrip human control.

The IPO question has followed OpenAI since its 2022 launch of ChatGPT turned it into one of the most valuable private companies in the world. A public listing would hand early investors and employees a path to liquidity, but it would also force quarterly disclosure of finances and strategy at a moment when OpenAI is spending heavily on compute and talent. Altman's comments suggest the board and leadership see more downside than upside in that trade-off right now.

Altman has previously floated the possibility of an eventual IPO without committing to a date, and his latest remarks push any listing well past the near term. The company's unusual corporate structure — a capped-profit arm overseen by a nonprofit board — would itself complicate a traditional offering and require significant legal work before bankers could be brought in. That structure has already drawn scrutiny from regulators and former employees.

The interview also touched on recursive self-improvement, the idea that an AI system could rewrite its own code and accelerate its own capabilities without human input. Altman has spoken about the concept before, and his willingness to discuss it in the same conversation as an IPO timeline reflects how OpenAI frames its business decisions around safety arguments. He also addressed the Hugging Face hacking incident, which raised questions across the industry about how exposed open model repositories and their users really are.

On the prospect of an AI system beyond human control, Altman did not rule it out. That position sits awkwardly beside a public offering, since a company selling shares must tell investors what it can and cannot govern. For now, OpenAI stays private, funded by large strategic partners rather than the open market.

What to watch next is whether Altman's stance holds if rivals begin listing. Anthropic and xAI remain private, but a successful AI IPO elsewhere could change the calculus fast, and OpenAI's investors will want a timeline eventually. The other open question is whether the capped-profit structure survives intact if a listing ever moves from hypothetical to planned.