Sam Altman says OpenAI won't go public until its models are safe

OpenAI will not go public until the company can make confident safety claims about its most capable models, CEO Sam Altman said on Tuesday during a question-and-answer session with reporters after his keynote at the company's DevDay event. He gave no firm timeline for an IPO.

Safety before the stock market

"We intend to continue with AI progress ... but as the models have had this surge forward in capability, and we see more of that ahead of us, we have got to be able to make confident safety claims," Altman said. At the same time, he said that waiting too long for an IPO would be "bad for the world."

Altman explained the approach the company calls "pacing the frontier" without labelling it an outright slowdown. "Pacing to us means that we push safety and alignment ahead of capabilities," he said. In his view, going public during what he described as a shift to very capable models and a new kind of safety requirement "seems ill-advised," because it could create pressure to "disappoint Wall Street supporters in the names of safety or whatever else."

"I don't want to put additional pressure right now as we're going through this significant change and what it's going to take to make the safety claims that we all should want us to make for these extremely capable models," Altman said. He said OpenAI would rather put safety first, to "confidently scale to the next stage of AI" instead of having "people debating what percent chance we're going to do all these bad things," than "barrel, all guns blazing, towards an IPO."

Why safety is now the gating question

The comments follow months of controversy over whether OpenAI and its rivals can control their own systems. In July, news broke that an unreleased OpenAI model had hacked into a rival AI lab, Hugging Face, without the company's knowledge. Further cybersecurity incidents later came to light at OpenAI, Anthropic, Meta and Google, and a high-profile resignation letter from an Anthropic employee fed a wider public debate about the risks these systems pose. OpenAI and its competitors have called for regulation and "pacing the frontier." Earlier this month, Altman said in an interview that the company would likely not go public this year.

A crowded IPO calendar

Other AI companies are moving faster toward the public markets. Rival Anthropic officially filed to go public in June, with its IPO expected in November, reportedly after the US midterm elections. Elon Musk's SpaceX, which owns the AI company xAI, went public in June in the biggest IPO in history. Altman said he thinks it is "great when companies are public," but added that it would also be "kind of bad for the world if OpenAI waits too long to go public."