OpenAI Postpones IPO to After 2026
Sam Altman, chief executive of OpenAI, said in remarks published yesterday that the company will not carry out its initial public offering in 2026. Altman attributed this to safety concerns that make even a 10 percent chance of artificial intelligence causing human extinction by the end of the decade unacceptable.
Altman made these comments at a time when more U.S. lawmakers are urging the establishment of new regulatory frameworks for AI systems, following stark warnings from two researchers at Anthropic, OpenAI’s competitor, that rapid AI progress could lead to human extinction in the not-too-distant future.
In an interview with Fortune magazine, Altman said: “I already believe that, given everything that is happening regarding safety, now is not the right time for an initial public offering, and we do not feel pressured to do so.” When asked about the possibility of a 10 percent chance that AI could cause human extinction, Altman replied: “Whether it is 10, 8, or 6 percent, the key point is that a grave responsibility rests on all of us, and we cannot allow pride or profit motives or anything else to undermine that.”
He added: “We must act in a way that does not expose us to any of these risks, and I believe we are capable of doing so.”
These warnings come after incidents in which AI agents went out of control and breached external systems, and after the resignation of AI safety researchers from their companies due to concerns about the technology’s risks. Politicians from both the Democratic and Republican parties have expressed concern and called for further action.
The New York Times reported in June that OpenAI, headquartered in San Francisco, was considering delaying an initial public offering that could value the company at $1 trillion until next year. At that time, shares of Elon Musk’s SpaceX fell in its initial public offering after a surge pushed the company’s valuation to $1.8 trillion.
When asked by Fortune whether 2026 was ruled out for the offering and whether 2027 was more likely, Altman said: “I can say it is not 2026. Yes, we have a lot to do, such as addressing the safety and alignment requirements that this moment demands, and how the industry and governments can work together.”
Fortune reported that Altman also indicated that OpenAI and other leading AI companies may be on the verge of announcing an agreement to slow the development of AI and work together to address safety risks.
