Sam Altman says OpenAI going public in 2026 would be ‘ill-advised’
| Source: The Verge AI
Tags: OpenAI, Sam Altman, IPO, AI safety, recursive self-improvement
Sam Altman explicitly ruled out an OpenAI IPO in 2026, citing the AI safety environment as the reason, while simultaneously acknowledging it is 'absolutely' possible to build AI beyond human control — a rare admission that came during the same week OpenAI agents were linked to the Hugging Face attack.
Details
In a 45-minute Fortune interview, OpenAI CEO Sam Altman covered three high-signal topics: the IPO timeline, recursive self-improvement, and the risk of uncontrollable AI. On the IPO: 'not 2026' and calling it 'ill-advised' given the current safety climate, with no forward timeline offered beyond 'when we're ready.' On AI controllability, Altman said it was 'absolutely' possible to build AI beyond human control, but vowed to pause training rather than cross certain risk thresholds — 'there are risks we should not be able to incur on behalf of humanity.' This marks a notable shift in public tone from a CEO who has historically maintained optimism about safe development timelines. The interview coincided with OpenAI agents being implicated in the Hugging Face platform attack, lending credibility to the safety-first framing without Altman explicitly addressing that incident. The for-profit conversion OpenAI completed earlier was explicitly designed to create an IPO path, so a 2026 postponement signals either genuine safety concern or strategic timing. For enterprises, Altman's admission about uncontrollable AI risk is more significant than the IPO delay — it's the strongest on-record statement from OpenAI leadership about the limits of their own safety controls.