More than 1,386 employees from frontier AI companies have signed a public statement calling for international coordination to slow the pace of AI development as systems approach human-level capabilities.

The signatories include senior figures from OpenAI, Anthropic, Google DeepMind, and Meta AI. Notable names include OpenAI Chief Scientist Jakub Pachocki, Anthropic co-founder Jared Kaplan, and Safe Superintelligence CEO Ilya Sutskever.

The statement warns that leading AI companies "believe they could be close to automating AI research" and that capability development risks accelerating "beyond our ability to understand or control the resulting systems."

Why coordination matters now

Employees cite mounting evidence of rapid progress. Dawn Song, VP of AI Research at Meta, noted that frontier AI agents can now "discover and exploit real-world software vulnerabilities" at scale through systems like CyberGym and ExploitGym.

Shengjia Zhao, Chief Scientist at Meta AI, said frontier labs are "very close to AI that can exceed even the best people on almost every metric of intelligence."

The statement acknowledges that individual companies and countries face "intense competitive pressure not to unilaterally slow that acceleration" despite potential risks.

Technical governance gaps

The group requests US government support for developing "technical and governance tools needed to deliberately pace the frontier of automated AI development."

John Schulman from Thinking Machines Lab said he wants to see "labs start designing these mechanisms voluntarily, even before the USG gets involved."

Matthew Rahtz, a Staff Research Engineer at Google, described the recent pace as "a shock" even after three years working on AI capability evaluations.

The statement builds on existing work to monitor frontier model releases but calls for broader international coordination mechanisms. Signatories emphasize the need for tools to "buy time to address emerging risks, develop security measures, and strengthen oversight" before automated AI research accelerates further.