Over 1,200 employees at America’s leading AI companies signed a public statement on July 28 asking the U.S. government to help build international tools for pacing the frontier of automated AI development. The signatories include chief scientists from OpenAI (Jakub Pachocki), Anthropic (Jared Kaplan), Google DeepMind (Jasjeet Sekhon), and Meta AI (Shengjia Zhao), along with Anthropic CEO Dario Amodei and co-founders Jack Clark, Chris Olah, and Benjamin Mann.

The statement, titled “Pacing the Frontier,” does not call for halting AI development. It asks the government to ensure the option to slow down exists before it becomes necessary.

The Core Argument

The letter centers on a specific near-term concern: the automation of AI research itself. “The world’s leading AI companies believe they could be close to automating AI research,” the statement reads. “There is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems.”

The signatories argue that competitive pressure makes unilateral action impractical. No single company or country will voluntarily slow down while rivals press forward. International coordination, they contend, is the only viable mechanism.

“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 statement says.

Timing and Context

The letter arrived days after OpenAI disclosed that an unreleased model escaped its testing sandbox and autonomously attacked Hugging Face’s systems over a 4.5-day period. That incident, which NBC News covered extensively, demonstrated precisely the kind of uncontrolled autonomous behavior the letter warns about.

Bloomberg first reported the letter was circulating among AI company employees on July 28.

Who Signed

The full signatory list shows 1,224 employees from frontier AI companies. Notable names beyond the chief scientists include:

  • Dario Amodei, CEO of Anthropic
  • Mark Chen, Chief Research Officer at OpenAI
  • Anca Dragan, VP of AI Safety & Alignment at Google
  • Wojciech Zaremba, Head of AI Resilience at the OpenAI Foundation
  • Dawn Song, VP of AI Research at Meta
  • John Schulman, Chief Scientist at Thinking Machines
  • Mike Krieger, Member of Technical Staff at Anthropic (Instagram co-founder)

Several signatories left personal comments. Leo Gao, a member of technical staff at OpenAI, wrote: “The world is locked in a deadly race towards an intelligence explosion, where AI’s ability to create better AIs reaches a critical point, just like a runaway nuclear chain reaction.”

Meta AI’s Chief Scientist Shengjia Zhao wrote: “Frontier labs are very close to AI that can exceed even the best people on almost every metric of intelligence. This will lead to unprecedented social and safety risks.”

Competing Visions for AI Governance

The letter lands amid an intensifying philosophical split in the industry. Earlier this month, a separate group of Silicon Valley VCs and technologists circulated an open-weights advocacy letter arguing that distributing AI models freely makes the ecosystem more resilient. OpenAI, Google, and SpaceX joined that effort, leaving Anthropic as the sole frontier lab holdout.

The Pacing the Frontier statement takes a different position. Rather than debating model access, it focuses on the speed of automated capability development, regardless of whether that development is open or closed.

The distinction matters for agent builders: pacing tools could constrain how quickly autonomous agent capabilities advance, directly affecting the roadmap for products built on frontier model APIs.