OpenAI delays its 2026 IPO, citing unfinished safety work
Sam Altman said OpenAI will not go public in 2026 and that a possible listing would occur no earlier than 2027. Bloomberg Law and DIE ZEIT report that he considers the timing unsuitable amid the current safety debate and says work remains on safety, alignment and cooperation with governments.
Artificial Intelligence··Midday
The 2026 option comes off the table
Sam Altman told Fortune that OpenAI will not conduct an initial public offering in 2026. Bloomberg Law reports that he moved a possible listing to 2027, while DIE ZEIT says he considers the present environment unsuitable for going public. The statement changes expectations for a closely watched transaction. Neither report identifies a regulatory filing, fixed sale date or pricing. Thus 2027 is the earliest stated possibility, not a completed timetable.[1], [2]
Altman puts safety work first
Altman says OpenAI has work to do on safety and alignment, as well as on how industry and governments should act together. DIE ZEIT emphasizes the link between the decision and debate over controlling highly advanced AI. That rationale does not establish a conclusion about the company's finances: the reports provide no revenue, cash-need or valuation data. The narrower supported decision is that the chief executive places safety work ahead of the current listing timetable.[1], [2]
The private-company structure remains for now
Delaying the offering means OpenAI remains private during this period. Altman says the company feels no pressure to list immediately. His position implies that safety decisions can be addressed outside public markets' periodic expectations, but the reports do not independently measure whether private ownership produces better safety. Concrete indicators for any revival of the transaction would be a filed prospectus, an announced timetable and the criteria by which OpenAI considers its safety work sufficiently complete.[1], [2]
