Altman Rules Out OpenAI IPO Until It Can Make Firmer Safety Promises

OpenAI chief executive Sam Altman says the company will stay private until it can offer stronger guarantees about the safety of its models, with no timeline attached.
OpenAI will not pursue an initial public offering until it can make better commitments about model safety, chief executive Sam Altman said, putting a hold on any near-term Wall Street debut for the most valuable private company in artificial intelligence. Altman gave no date or conditions that would trigger a listing, leaving the question of when OpenAI shares might trade publicly open. The comments, reported by The Verge, close off for now one of the most anticipated tech offerings in years.
Altman tied the decision directly to how quickly the company's models have improved. He said OpenAI intends to keep pushing AI progress, but the jump in capability has made it harder to promise investors and the public what those systems will and will not do. That gap between what the company can build and what it can guarantee is now the stated reason for staying private, rather than a conventional calculation about market conditions or valuation.
Going public would put OpenAI under quarterly reporting requirements and securities-disclosure rules that sit awkwardly with a business whose core product changes faster than a fiscal calendar. A public company must also tell shareholders in advance about material risks, and frontier-model behavior is difficult to characterize that way. Altman's framing suggests he sees those obligations as premature rather than merely inconvenient.
OpenAI already operates under a hybrid structure, with a nonprofit parent and a capped-profit arm, and it has raised billions from Microsoft and other backers without selling stock to the public. That arrangement gives it access to capital while avoiding the disclosure regime an IPO would impose. Rivals including Anthropic and Elon Musk's xAI remain private as well, so the sector's largest labs are not yet tested by public-market scrutiny.
The safety condition Altman attached is itself undefined. He did not say which benchmarks, audits or governance changes would count as sufficient, nor who would judge them. Until those standards exist, the IPO question rests on an internal call rather than an external milestone. Watch whether OpenAI publishes concrete safety criteria, and whether investors keep funding the company on private terms for as long as that takes.