OpenAI will not go public in 2026, Chief Executive Sam Altman said in a new interview, pushing one of Wall Street’s most anticipated listings into at least next year. Altman told Fortune that an initial public offering would be poorly timed while the industry confronts increasingly serious questions about the safety and control of advanced artificial intelligence. He said OpenAI does not feel financial pressure to list immediately and would be willing to slow or stop development if the company concluded that a system could not be built safely. The comments turn what had been a shifting timetable into the clearest public statement yet that a 2026 debut is off the table.
The decision separates OpenAI’s access to capital from the discipline and disclosure that come with public markets. The company has raised enormous sums privately to pay for chips, data centers and research, giving management more flexibility over timing than most businesses approaching an IPO. Reports that executives had already signaled a delay internally made the timing less surprising, but Altman’s direct link to safety concerns adds a new rationale and a public commitment. At the same time, a listing had been expected to test whether investors would support a valuation above $1 trillion for a company with extraordinary growth prospects and equally extraordinary infrastructure costs. Altman’s explanation puts safety governance, cooperation with governments and potential industry coordination ahead of that near-term valuation event.
Prediction markets repriced the upper end of OpenAI’s private valuation after the remarks. The Polymarket contract asking whether the company’s valuation will reach at least $1.25 trillion by year-end fell 19 percentage points over twenty-four hours to 40.5%. The $1 trillion threshold dropped 13.5 points to 69%, while the $1.5 trillion contract declined 10.5 points to 28.5%. The full family had about $29,600 in activity during the latest day. Those contracts concern the highest reported valuation before December 31, not whether an IPO occurs, so a private financing or secondary transaction could still move them.
The delay also changes the competitive calendar for the artificial-intelligence sector. Anthropic has continued preparing for a possible offering, while OpenAI can now focus on internal safeguards and any wider agreement among leading labs without simultaneously marketing shares. Investors will look for evidence that the pause produces concrete governance changes rather than an indefinite postponement. The next meaningful signals include any formal safety pact, additional disclosure about how OpenAI evaluates frontier systems, and new financing terms that reveal what private investors will pay. Any eventual prospectus will also have to explain how massive infrastructure commitments and safety obligations fit together. Until then, Altman has made the priority clear: the company’s public-market milestone will wait while OpenAI argues that the industry must first prove it can manage the risks created by its most capable models.
