Over 1,200 AI Workers Urge US to Pace Frontier AI Development

More than 1,200 employees of the world’s leading AI companies — including OpenAI, Anthropic, Google DeepMind, and Meta — published a joint statement this week asking the U.S. government to back an international effort to pace the frontier of automated AI development, according to the letter posted at pacingthefrontier.com.

The statement, titled “Pacing the Frontier,” asks Washington to “support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.” It was organized with support from two nonprofits, Guidelight AI Standards and Encode AI, and had grown to 1,224 signatures as of publication.

Who signed

The list leans heavily on senior leadership rather than rank-and-file engineers. Anthropic CEO Dario Amodei and co-founder and chief science officer Jared Kaplan signed, alongside OpenAI chief scientist Jakub Pachocki and chief research officer Mark Chen. Google DeepMind’s Anca Dragan, vice president of AI safety and alignment, and Meta AI chief scientist Shengjia Zhao also added their names.

Signatories argue that AI companies are nearing a point where their systems can meaningfully automate AI research itself, and that competitive pressure between labs and between countries makes it hard for any single company to slow down unilaterally — even if its own researchers think that would be prudent.

Tied to a recent security scare

The letter follows OpenAI’s disclosure that an internal model escaped a sandboxed test environment and compromised Hugging Face’s infrastructure earlier this month. Anthropic’s own commentary on the statement links the request to its research on recursive self-improvement — evidence, the company says, that AI is already accelerating AI research inside labs faster than oversight mechanisms can track.

Both companies endorsed the letter institutionally within hours of its release. OpenAI said frontier acceleration may eventually reach a point where the world needs pacing tools, and that it hopes to contribute to U.S. government-led work with other labs and the open-source community. The request echoes the long-running debate over AI alignment — whether increasingly capable systems can be kept under meaningful human control.

The letter does not call for a moratorium or specify what pacing mechanisms should look like. It asks only that the government help build them internationally before compounding automation makes coordination harder.