OpenAI IPO pushed to 2027 as Altman cites AI safety concerns

OpenAI will not go public in 2026, CEO Sam Altman confirmed in an interview published Saturday by Fortune, pushing back what would be one of Silicon Valley's most closely watched initial public offerings.

The San Francisco-based company filed confidentially for an IPO in June but left the timing open. Altman's comments now rule out a 2026 listing as the artificial intelligence industry faces mounting questions about how quickly developers should scale their most powerful systems.

"Right now would be an ill-advised moment to go public," Altman told Fortune. When asked directly whether that meant no IPO in 2026, he replied, "I would say not 2026. Yeah, we got a lot of stuff to do."

Safety work takes priority

Altman pointed to ongoing work on safety and alignment — the technical challenge of ensuring advanced AI systems behave as intended and remain under human control — as the primary reason for the delay.

"We got a lot of stuff to do, and like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together," he said.

The CEO disclosed that OpenAI has discussed pausing training runs when models reach new capability thresholds to allow researchers more time to develop safeguards. "We've talked in the last couple of months about pausing runs as we get to these new level of capabilities to make more safety and alignment progress," Altman said. "We will do more of that going forward."

Altman also said the company should not train models "where we cannot make a safety case for why we will be able to make strong statements about their controllability and alignment."

Industry-wide debate intensifies

The announcement follows a week of heightened debate within the AI industry about development pace. Anthropic CEO Dario Amodei called Saturday for AI companies to slow improvements to their most powerful models. Altman said he agreed with the need to "pace the frontier."

OpenAI's chief scientist, Jakub Pachocki, has also recently advocated for stronger standards, monitoring, and third-party oversight as AI systems advance in capability.

Why it matters

The IPO delay removes near-term pressure to prioritize quarterly results over safety research at a moment when frontier AI labs are grappling with systems that may be approaching unpredictable capability jumps. For San Francisco, it postpones a wealth-creation event that could have accelerated already-tight housing and office markets. The city has seen more than 140 homes sell for at least $1 million over asking in the first half of the year, and rental vacancy rates have dropped to 2.2%, the lowest among major U.S. cities, according to Apartment List. However, much AI-generated wealth is already circulating through high salaries and private stock sales, and San Francisco lacks local income or capital gains taxes that would directly capture IPO windfalls.

Altman did not rule out a public offering at a later date. "We need to be able to make decisions that are not obviously in the interest of our business and our shareholders," he said, "for the responsibility of fulfilling our mission and what that's going to require."

The details were first reported by the San Francisco Chronicle.