Altman says OpenAI is open to slowing frontier AI development

OpenAI is considering slowing down the development of cutting-edge artificial intelligence, chief executive Sam Altman told employees^2,4. Altman said he hopes competing firms adopt similar slowdowns⁵, and Bloomberg reported he hopes other companies will do the same¹. Quartz and Firstpost carried the same account of the staff meeting^6,7.

The New York Post reported Altman’s openness to slowing as safety risks mount, and referenced a role for government in helping to deliberately pace development³. PYMNTS characterised the remarks as floating an industrywide pause as frontier safety concerns grow⁴. The reporting is secondhand throughout: these are accounts of an internal meeting rather than a public commitment¹⁰.

Why it matters to investors

The pace of frontier development is an input to capital spending plans across the sector, and a lab conceding on pace is a different regulatory assumption from one resisting it. OpenAI has also reversed its position on rules: the company made a U-turn and called for binding national AI safety rules⁹, a shift Tech Xplore described as OpenAI calling for US rules on powerful AI⁸.

That reversal followed the release the company’s president, Greg Brockman, characterised as the beginning of the era of artificial general intelligence^8,9. A company that ships a model it calls a threshold moment and then asks for binding federal rules is describing a different cost structure for the industry than the one priced in a year ago.

Washington is moving in parallel. A Republican-led Senate subcommittee that oversees disaster management is investigating OpenAI’s handling of the Hugging Face breach in July¹¹. Senator Josh Hawley wrote to Altman noting that more AI experts are warning about existential risks, and cited three Anthropic researchers who said publicly there is a greater than 10% chance AI could kill all humans¹¹.

What to watch

The nearest financial consequence runs through the IPO pipeline. White House AI and crypto czar David Sacks is calling for Anthropic’s planned initial public offering to be paused after a former company researcher’s warning¹². Sacks, who was the administration’s crypto and AI adviser until March, used Altman’s remarks to question that listing directly¹⁰.

Two things would turn this from rhetoric into something investors can price. The first is whether any slowdown is actually implemented rather than floated, and whether rivals follow as Altman says he hopes^1,5. The second is whether the binding national rules OpenAI now says it wants⁹ arrive in a form that binds its competitors equally. Until then, the concrete items are the Senate investigation¹¹ and the pressure on the Anthropic listing¹².