OpenAI IPO Will Not Happen in 2026, Altman Says, Citing AI Safety Fears

LOS ANGELES, Sept 12 (Reuters) - In a Fortune interview published Saturday, OpenAI CEO Sam Altman said the company will not go public in 2026, citing concerns over artificial intelligence safety.

Altman told Fortune that with everything unfolding around safety, now would be an unwise time for an IPO, and that OpenAI feels no pressure to pursue one.

After two researchers at Anthropic, an OpenAI competitor, warned that fast-advancing AI could cause human extinction in the not-too-distant future, a growing number of U.S. lawmakers have demanded new rules for AI systems.

Those warnings came amid instances of AI agents turning rogue and hacking external systems, as well as AI safety researchers leaving their companies over concerns about the technology's dangers. Democrats and Republicans in politics have reacted with alarm and calls for further action.

In June, The New York Times reported that San Francisco-based OpenAI was weighing whether to delay a potential trillion-dollar IPO until the next year. At that time, shares in Elon Musk's SpaceX IPO were declining after a surge that had lifted the company's valuation to $1.8 trillion.

When Fortune asked whether 2026 was off the table in favor of 2027, Altman said it would not be 2026. He noted that the company had much to do, including meeting what this moment requires for safety and alignment and figuring out how industry and governments can work together.

According to Fortune, Altman also indicated that OpenAI and other top AI companies could be close to unveiling a pact to slow AI development and cooperate on safety risks.

On Saturday, Anthropic CEO Dario Amodei pressed AI companies to adopt a more deliberate approach to development.

In an essay shared on social media, Amodei wrote that AI companies must slow the pace at which they improve the capabilities of AI models.

Altman later posted that he shared that sentiment.

In a response to Amodei's post on X, Altman wrote that he agreed with Dario that the frontier needed to be paced. He added that this had been a primary topic of discussions at OpenAI in recent weeks.

So far, safety concerns have not slowed Anthropic's own IPO plans. People familiar with the matter told Reuters this month that Anthropic is expected to begin marketing its initial public offering in mid-October at the earliest and complete the listing days before the U.S. midterm elections in November.