Altman defers OpenAI's IPO past 2026 on safety grounds
Sam Altman told Fortune that a listing now would be ill-advised while safety and alignment work is unfinished. More than $180 billion has gone into OpenAI, on Fortune's count, with no scheduled exit.
The Investor · Invest desk
What happened
- Altman told Fortune editor-in-chief Alyson Shontell on Friday, in an interview for her Titans and Disruptors of Industry podcast, that it was not the right time for OpenAI to go public.
- He said OpenAI will list when the business is ready and when the company is ready as it relates to what the moment is like in society with this technology.
- The remarks followed researcher Jacob Coxon's public resignation from Anthropic, with a message that AI's makers are moving quickly and irresponsibly, and Dario Amodei's response over the weekend.
- Altman also suggested the leading AI figures, from Amodei to Hassabis, will likely get together on safety at some point, while declining to pre-announce private discussions.
- Investors have poured more than $180 billion into OpenAI over the years, by Fortune's count, and the much-anticipated exit appears far on the horizon.
Why it matters
- What doesn't add up contradiction Fortune's headline puts the window at 2027, while the only year in Altman's quoted answer is the 2026 he excluded, so a model built off the headline is a year more precise than the source it cites.
- What it costs cost Each percentage point of annual cost of capital on more than $180 billion runs about $1.8 billion a year, and the funds holding the position carry that, not OpenAI.
- What it limits constraint The condition Altman set is society's readiness for the technology, which no outside party scores, so anyone dating an OpenAI exit is forecasting the issuer's own judgement.
- What it sets up precedent Once the largest private AI company makes safety and alignment work a listing condition, the next lab to approach the public market will be asked to state its own version.
Next story · Invest · Sep 30
OpenAI's $1.4 trillion ask would mark up its biggest backers' stakes by about two thirdsConfirmed4 publishers
"We don't feel pressure on that" is the clause in Friday's answer with financial consequence [2]. A company that has taken more than $180 billion from private investors does not need a public market to fund itself [7]. The same newsletter that carried Altman's remarks listed $875 million going into Positron AI [11], $550 million into Temporal [12], $205 million into Cornelis [13] and $130 million into Buildots [14]. That is $1.76 billion into four private companies in a single day's deal section [15], or close to 1% of everything OpenAI has raised [16].
His own words on timing were narrow. "I would say not 2026," Altman said when pressed on whether 2026 was off the table in favour of 2027. "Yeah, we got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together" [3].
Two readings fit that sentence. If the safety concern is binding, a filing waits on an industry-and-government arrangement nobody has built yet, and the wait is open-ended. If it describes the mood of one week in September, the condition is loose enough that a friendlier market next year clears it with nothing changed in alignment research.
In my view the framing is accurate and non-binding at the same time. It is the sentence available to a company with no funding need. Fortune's Allie Garfinkle, who reported the interview, wrote in the same newsletter: "I do wonder if it's easier to talk about apocalyptic fears of the future than concerns of the present" [8]. Coxon and Amodei made the counter-argument under their own names [6], and if that argument produces something written down, the 2027 question stops being about market timing. What would prove me wrong is OpenAI publishing a readiness standard an outsider can check, with dates in it.
What to watch
- An OpenAI S-1, public or confidential, before the end of 2026 would show the safety condition was not binding.
- A tender or secondary sale that gives early OpenAI holders liquidity without a listing.
Information, not investment advice. This story is general information about companies and markets, drawn from cited reporting. It is not a recommendation to buy, sell or hold any security; check the cited sources before acting.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence62
- Adoption45
- Hype gap+28
- Incentives72
- Confidence66
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
On Friday, OpenAI CEO Sam Altman told Fortune editor-in-chief Alyson Shontell, in an interview for her Titans and Disruptors of Industry podcast, that it wasn't the right time for OpenAI to go public.
- [2]
"I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don't feel pressure on that," Altman said.
- [3]
Pressed on whether 2026 is off the table in favor of 2027, Altman replied: "I would say not 2026. Yeah, we got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together."
- [4]
Altman added that OpenAI will go public when the business is ready and when the company is ready as it relates to "what the moment is like in society with this technology."
- [5]
Fortune's headline on the interview said Altman says OpenAI's IPO window has been pushed to 2027, but markets aren't the culprit.
- [6]
Researcher Jacob Coxon, who had also worked at OpenAI, resigned from Anthropic last week with a very public message that AI's makers are moving quickly and irresponsibly, and Anthropic CEO Dario Amodei sounded off over the weekend.
- [7]
Investors have poured more than $180 billion into OpenAI over the years, and a much-anticipated exit appears far on the horizon.
- [8]
Fortune's Allie Garfinkle wrote: "I do wonder if it's easier to talk about apocalyptic fears of the future than concerns of the present."
- [9]
Altman suggested that AI's biggest names, from Amodei to Hassabis, will likely get together to discuss safety at some point, while telling Shontell he would not pre-announce private discussions.
- [10]
The only calendar year Altman named in his quoted answer is 2026, which he excluded; the 2027 date appears in Fortune's headline rather than in his quoted words.
- [11]
Positron AI, a Reno, Nevada-based developer of AI inference hardware and software, raised $875 million in Series C funding from NEA, Atreides Management, Valor Equity Partners, Andra Capital and others.
- [12]
Temporal, a Bellevue, Washington-based developer of software for building reliable AI applications, raised $550 million in Series E funding led by Lightspeed, Wellington Management, Growth Equity at Goldman Sachs Alternatives and Tiger Global.
- [13]
Cornelis, a Wayne, Pennsylvania-based developer of high-performance networking technology, raised $205 million in funding led by IAG Capital Partners.
- [14]
Buildots, a Tel Aviv-based developer of AI software for construction projects, raised $130 million in funding led by O.G. Venture Partners.
- [15]
The four largest venture rounds in the same day's Fortune deal section total $1.76 billion.
- [16]
That $1.76 billion is about 0.98% of the more than $180 billion investors have put into OpenAI.
- [17]
One percentage point of annual cost of capital on $180 billion is $1.8 billion a year.
Sources
2 independent publishers whose own reporting we read for this story.
- cryptobriefing.comOpenAI and Anthropic IPOs face growing uncertainty as AI development slowdown reshapes investor expectations
1 article · September 15, 2026
- fortune.comSam Altman says OpenAI’s IPO window has been pushed to 2027—but markets aren’t the culprit
2 articles · September 15, 2026