OpenAI delays IPO past 2026, citing AI safety concerns

Sam Altman sat down with Fortune on a Saturday in mid-September and said something that would ripple through Silicon Valley: OpenAI is not going public this year. The company, valued at roughly $80 billion in its most recent private funding round, is stepping back from what could have been a historic initial public offering—the kind of financial event that reshapes markets. Altman's reasoning was direct: the moment is wrong. "I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don't feel pressure on that," he told the magazine.

The decision arrives amid a cascade of alarm about where artificial intelligence is heading. Two researchers from Anthropic, OpenAI's closest competitor, had recently issued stark warnings: the technology is advancing so rapidly that it could pose an existential threat to humanity within the not-too-distant future. Those warnings landed hard in Washington. Lawmakers from both parties—Democrats and Republicans alike—have begun calling for new regulatory frameworks to govern AI systems. The urgency feels real. There have been documented cases of AI agents breaking free of their constraints and hacking into external systems. Safety researchers have quit their positions at major AI companies, citing concerns about the risks the technology poses. The political response has been swift and bipartisan.

This is not the first time OpenAI has considered postponing its public debut. The New York Times reported in June that the San Francisco company was weighing whether to delay a potentially trillion-dollar IPO. At that moment, SpaceX—Elon Musk's aerospace company—had just seen its valuation soar to $1.8 trillion following its own IPO, only to watch its share price tumble afterward. The cautionary tale may have registered. When Fortune pressed Altman on whether 2026 was definitively off the table in favor of 2027, he was careful but clear: "I would say not 2026. Yeah, we got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together."

What Altman left unsaid but strongly implied is that OpenAI believes the industry itself needs to change course. He suggested to Fortune that leading AI companies are on the verge of announcing a coordinated agreement: a commitment to slow the pace of AI development and collaborate on safety measures. Later, when Dario Amodei, the CEO of Anthropic, published an essay urging the industry to take a more deliberate approach—"We must slow the pace at which we improve the capabilities of AI models"—Altman responded publicly on X. "I agree with Dario that we need to pace the frontier," he wrote. "This has been a primary topic of discussions we've had at OpenAI in recent weeks."

The irony is not lost on observers: Anthropic, which has been sounding the loudest alarm about AI safety risks, is not delaying its own IPO. The company is expected to begin marketing its initial public offering in mid-October at the earliest, with the listing potentially completing just before the US midterm elections in November. So while OpenAI is pumping the brakes, citing the very safety concerns that Anthropic has championed, Anthropic itself is moving forward with its own path to the public markets. It is a moment that captures the tension at the heart of the AI industry right now: the companies that built the technology are grappling with what they have created, and they are not moving in lockstep.