Story
September 14, 2026
OpenAI Puts Its Trillion-Dollar IPO on Hold as Safety Fears Surge
OpenAI frames its IPO delay as room to put safety and public responsibility ahead of investor demands, while the broader AI debate has split between those urging a slower frontier and critics who see safety rhetoric as a strategic maneuver.
OpenAI had been widely expected to stage one of Silicon Valley’s biggest-ever market debuts. Instead, the company’s chief executive, Sam Altman, has made clear that Wall Street will have to wait.
In June, OpenAI confidentially filed for a potential IPO, but never announced a target date.1 The prospect of a public listing — potentially valuing the company in the trillions — arrived as its models became more capable and the argument over who can safely control them grew sharper.
That tension intensified in the days before Altman’s latest comments. Former OpenAI and Anthropic researcher Jacob Coxon left Anthropic, warning on X that people building AI “earnestly believe that it could kill us all by the end of the decade.”2 Anthropic chief Dario Amodei then called for the industry to slow the improvement of its most powerful systems: “We must slow the pace at which we improve the capabilities of AI models.”3
In a Fortune interview released Saturday, Altman answered with an unusually direct timetable: no OpenAI IPO in 2026. “Given everything happening with safety, right now would be an ill-advised moment to go public,” he said, adding that the company had “a lot of stuff to do” on safety, alignment and cooperation with governments.4
Altman’s case is not merely that caution is good public relations. He said OpenAI may pause training at new capability thresholds and argued it should not train systems without a credible safety case for their “controllability and alignment.”5 Staying private, in this view, protects the company’s ability to make decisions that are not plainly in the interests of the business or its shareholders.6
The emerging consensus on pacing is hardly universal. Altman publicly agreed that “We need to pace the frontier,”
7 but Yann LeCun amplified a critique portraying Amodei’s proposal as an IPO-era branding exercise wrapped in regulatory capture.
8 David Sacks, meanwhile, amplified the argument that firms warning AI is dangerous must reconcile that claim with asking public investors to fund trillion-dollar valuations.9
For now, Altman is betting that delaying a blockbuster listing is the more defensible choice than asking markets to finance a race whose guardrails remain unsettled.