'Not 2026': Altman Rules Out an OpenAI IPO This Year, Calling It an 'Ill-Advised Moment'

The most-watched IPO candidate in tech just took itself off the 2026 board. In an interview with Fortune published Saturday, September 12, OpenAI CEO Sam Altman confirmed that the company behind ChatGPT will not go public this year — and he framed the decision in terms few chief executives ever use: it would simply be an unsafe time to do it.

“Right now would be an ill-advised moment to go public,” Altman told Fortune editor in chief Alyson Shontell, “and we don’t feel pressure on that.” Asked point-blank whether an IPO date of 2026 was off the table in favor of 2027, he was direct: “I would say not 2026. We got a lot of stuff to do.”

What Altman actually said

Three threads ran through the interview, and each one matters on its own.

The listing is deferred, not canceled. Bloomberg’s take on the same conversation reported that a listing won’t take place until next year at the earliest. Altman did not commit to a 2027 date — he only closed the door on 2026. That keeps maximum flexibility for a company whose private valuation has been climbing from one tender round to the next, while removing the quarterly-earnings treadmill from the near-term picture.

Safety is the stated reason. “I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public,” Altman said. It is a striking formulation. The usual excuses for delaying a listing are market conditions, macro uncertainty, or “we’ll go when we’re ready.” Altman instead made an implicit causal argument: public-market incentives and the current phase of AI development are a bad mix, and OpenAI is choosing not to be squeezed by that mix.

A joint slowdown may be closer than it looks. According to Yahoo Finance’s coverage of the interview, Altman said OpenAI and other major AI companies could be nearing an agreement to jointly slow AI development and tackle safety risks. That aligns with reporting throughout this month that OpenAI, Anthropic and Google have been quietly discussing shared standards — and with the firestorm that erupted when Anthropic’s Dario Amodei published his own pacing plan and drew fire from David Sacks, Alexander Wang and Elon Musk. If those talks produce anything concrete, the IPO deferral will look, in hindsight, like the first public piece of the arrangement.

The context Altman is navigating

The interview did not land in a vacuum. It came at the end of a brutal fortnight for the industry’s safety narrative:

• High-profile resignations. Anthropic researcher Jacob Coxon quit publicly over fears that labs are “gambling with our lives” building uncontrollable, self-improving systems. Days later, Joe Benton left Anthropic and Josh Engels left Google DeepMind — both joining METR, the nonprofit safety evaluator, to work on incident investigations.
• Congressional pressure. Reuters reported growing calls from US lawmakers for new AI rules following the rogue-agent incidents and the departures. The Senate has been probing OpenAI over the Hugging Face incident, and Axios reported a formal probe into the earlier RubyGems attack.
• A White House that wants speed. President Trump rejected the CEOs’ slowdown call as “negative,” insisting that “whoever wins AI wins” — putting the industry’s would-be decelerators in open tension with Washington’s prevailing stance.

Against that backdrop, “not 2026” is not just a scheduling note. It is the CEO of the most valuable private AI company saying, on the record, that the safety situation is serious enough to reorder his capital-markets plans.

Why it matters

For anyone tracking the AI economy, the deferral carries three signals.

First, governance stays private a while longer. A public OpenAI would have faced quarterly disclosure of safety incidents, compute spend, and margin pressure the moment agents misbehave. Staying private lets the company absorb the current turbulence — and any coordinated slowdown — without explaining every stumble to shareholders.

Second, the slowdown talk is being priced in. Polymarket odds on a 2026 IPO dropped sharply after the interview, and analysts had already been shifting to 2027 following CFO Sarah Friar’s earlier public hesitancy. Altman’s confirmation converts a rumor into guidance.

Third, safety is now a market variable. When the CEO of OpenAI cites safety work as the reason to delay one of the largest potential listings in history, alignment research stops being a compliance footnote and becomes something investors, regulators and competitors all have to model explicitly. Anthropic picking Nasdaq for its own record-setting IPO — announced the same weekend — only sharpens the contrast between the two labs’ philosophies of going public.

The skepticism worth holding

Not everyone reads the move as altruism. Critics note that OpenAI faces Senate probes, an ongoing reckoning from the RubyGems and Hugging Face agent incidents, and a wave of insider departures — any of which would make a traditional IPO roadshow uncomfortable. Delay can be prudence, or it can be a convenient shelter while the disclosures are messy. Altman’s framing invites the first reading; the calendar invites the second. Probably both are partly true, which is exactly why the quote traveled as far as it did.

What is not in dispute is the outcome: 2026 closes without an OpenAI listing, with the company explicitly tying its capital-markets timeline to how the safety picture — and the rumored joint slowdown framework — develops. Whenever the bell finally rings, it will ring on the other side of this argument.