More than 1,100 workers from leading AI firms have signed a petition calling on the U.S. government to support tools that "pace the frontier of automated AI development"2,3.
The signatories include the CEO of Anthropic and the chief scientists from OpenAI and Meta Superintelligence Lab. The letter asks for a technical and policy framework that a government or international body could use, if necessary, to deliberately pace AI development.
Bloomberg AI editor Seth Figments, discussing the petition, characterized the backstory as driven by "a growing awareness that AI might one day be able to kind of automate itself" — that researchers are building technology capable of conducting its own research and development, potentially creating runaway progress. Figments noted that the signatories are "not specifically calling to slow it down, but that seems to be the implication".
Separately, Jack Clark, one of Anthropic's co-founders, previously posted a lengthy call for a pause button or mechanism to pause AI development.
Altman has expressed support for the effort, a stance described as rare for the OpenAI leader1.
The petition arrived days after what Bloomberg's coverage described as "the first of now two different but related breaches from OpenAI's" technology, though the excerpt did not elaborate further on those incidents.
ANALYSIS The letter's framing around "pacing" rather than "pausing" represents a calibrated ask — broad enough to attract over a thousand signatories from competing labs, yet specific enough to call for concrete technical and policy mechanisms. The participation of senior figures from Anthropic, OpenAI, and Meta Superintelligence Lab on the same petition is notable given those organizations' competitive positions in frontier AI development.