There will be no OpenAI IPO, the company behind ChatGPT, in 2026. The announcement came from OpenAI CEO Sam Altman, who cited safety concerns tied to artificial intelligence that make «unacceptable» a 10% risk that AI could end up causing human extinction by the end of the decade.

Altman: “This isn’t the right moment” to go public

The delay of OpenAI’s stock market debut is already triggering an immediate reaction on markets around the world, dragging down tech shares and AI-linked stocks in particular.

In an interview with Fortune, Altman explained the decision this way: «I really believe that, given everything happening around safety, this isn’t the right moment» for the company to go public. He added that, for now, OpenAI feels «no pressure» to launch an IPO.

A 10% extinction risk Altman calls unacceptable

Asked whether there is really a 10% risk of human extinction because of AI, Altman said he didn’t know how such a forecast had been arrived at, but warned the threat is serious enough that AI companies and governments alike have a duty to act to prevent it: «Whether it’s 10%, or 8%, or 6%, the point is that we all carry an enormous responsibility, and we cannot let ego, or profit incentives, or anything else get in the way».

In a post published on X shortly after midnight on Monday, September 14, Altman wrote: «We welcome a federal regulatory framework that sets uniform safety requirements for frontier AI», adding that «no competitive pressure from within the United States can justify irresponsible behavior». He laid out two scenarios in which AI progress could go «very wrong»: losing control of the future to AI itself, or an excessive concentration of power in the hands of a single person or company.

Anthropic’s Dario Amodei calls for a slowdown

Altman wasn’t the only one to speak out. Anthropic CEO Dario Amodei has also called for a slowdown in the development of AI models, warning that if companies in the sector don’t hit the brakes, within six to twelve months AI could become capable of directing swarms of agents able to seize control of the internet.

One of the most prominent voices in the industry, Amodei laid out in a post on Anthropic’s site a plan to strengthen oversight of the sector, noting that Anthropic — unlike OpenAI — is still prepared to go ahead with its own stock market debut, and is already implementing part of that plan.

As Amodei put it: «If slowing down development bought us even one or two years before models reach critical capability thresholds, and we used that time to make progress on alignment, we could meaningfully reduce the risk of something going seriously wrong».

A researcher’s resignation reignites the debate

The sell-off follows an extinction-risk warning last week from Jacob Coxon, an Anthropic researcher who resigned and went public with the alarm, accusing AI companies of «playing with our lives» as the situation slips beyond their own control. Coxon’s warning was echoed by other industry figures, including Evan Hubinger, who wrote on X that he too believes the probability of human extinction within the next decade from AI exceeds 10%. Several other researchers at Anthropic and OpenAI have publicly called for the pace of development to slow down.

AI stocks sell off from Asia to Wall Street

In a session already weighed down by rising oil prices, tech shares — and AI-linked names above all — fell sharply across global markets as investors priced in fading enthusiasm for the AI trade.

In Asia, shares of South Korean chipmakers SK Hynix and Samsung Electronics closed down more than 6% and 4%, respectively. In Japan, SoftBank — one of OpenAI’s largest investors — plunged 10%.

In Europe, semiconductor and other AI-linked stocks also sold off heavily in early trading: Dutch chip-equipment giant ASML lost more than 4%, Nokia fell 5%, and Infineon dropped more than 6%. Shares in data-center infrastructure names Siemens Energy and Schneider Electric also declined.

The selling extended into U.S. premarket trading, hitting chipmakers Micron, Intel and Nvidia, along with hyperscalers Microsoft, Amazon and Alphabet.

Editor’s note

This article was originally published in Italian on money.it by Laura Naka Antonelli on September 14, 2026 as «Perché OpenAI rimanda l’IPO? Per Altman rischio estinzione inaccettabile, sell off su azioni AI». It has been translated and adapted for an international audience by the Money.it International desk.