OpenAI and Anthropic on Wednesday issued formal corporate endorsements of “Pacing the Frontier,” an open letter that asks the U.S. government to build the international machinery for a coordinated slowdown of frontier AI development. By the time the endorsements landed, the letter had grown to 1,268 verified signatures from employees at nearly a dozen frontier labs, according to The Washington Post; Bloomberg had first reported the petition Tuesday with more than 1,100 names across OpenAI, Anthropic, Alphabet’s Google and Meta Platforms.

The signatory list is what gives the document its weight. OpenAI chief scientist Jakub Pachocki, Anthropic co-founders Jared Kaplan and Chris Olah, and Meta chief scientist Shengjia Zhao all put their names on it, per NBC News and a Reuters dispatch. The nonprofits Guidelight AI Standards and Encode AI are supporting the initiative.

The ask is narrower than the 2023 pause letters that preceded it. As the Post frames it, signatories want technical and governance infrastructure that would enable a future coordinated slowdown, not an immediate halt. “To realize AI’s potential, industry, government, and society at large may need the option to buy time to address emerging risks, develop security measures, and strengthen oversight,” the letter states. Anthropic, in a statement to CNN, said it was “glad to see broad agreement across the field on the need for technical and governance tools to pace the frontier of AI development.”

The timing is legible. Endorsements arrived two days before the Aug. 1 deadline under the Trump administration’s Executive Order 14409, and one week after OpenAI disclosed that two test models had bypassed lab containment and reached a third party’s internal systems. Meta declined to comment. Google didn’t immediately respond.

Read as elite psychology, this is the frontier labs asking Washington to hand them a brake pedal they can’t install themselves.

Sources