More than 1,200 employees at leading AI companies have signed a joint letter to the US government. Under the title Pacing the Frontier, they ask for tools that would allow the pace of automated AI research to be steered deliberately. The text pointedly does not call for an immediate pause or slowdown.
The letter asks for tools, not for using them now
The letter makes a single concrete request of Washington: the government should support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development. The rationale rests on the companies’ own assessment that they may be close to automating AI research itself – the point at which AI systems take over much of the work of building their successors.
From that, the text argues, follows a real risk that capabilities advance faster than the resulting systems can be understood or controlled. It also names the industry’s core bind: competitive pressure keeps individual companies from slowing down unilaterally, and the world lacks the means to pace frontier-wide progress collectively. Which instruments those might be is left open; the letter asks for their development, not their immediate use. Organisational backing comes from two independent non-profits, Guidelight AI Standards and Encode AI.
Signatures reach into the executive floors
The list of signatories runs all the way up. Among those named are Anthropic CEO Dario Amodei along with co-founders Jared Kaplan and Jack Clark, OpenAI chief scientist Jakub Pachocki, Meta chief scientist Shengjia Zhao, and Anca Dragan, who leads AI safety at Google. John Schulman, chief scientist at Thinking Machines, has signed as well. Anthropic and OpenAI back the letter officially, according to The Next Web, which sets it apart from earlier appeals mounted by individual staff.
The count keeps climbing. The outlet recorded 1,134 signatories on July 28; a day later the campaign website listed 1,224. No independent check of the details exists, including whether individual signatories work at the companies named. The letter follows an earlier push by Anthropic for a verifiable pause mechanism. Notably, the text carries no commitment of its own: the companies promise nothing and instead ask the state for a framework. The organisations involved disclose nothing about how the campaign is funded.
The pushback comes from inside the industry
Objections come from several directions. The economist Christian Catalini asks why China would join such a coordination effort at all. Former Microsoft executive Steven Sinofsky points out that company leaders could stop development themselves at any time – many of them signed the letter without doing so.
The sharpest criticism comes from Meta CEO Mark Zuckerberg, whose own chief scientist is among the signatories. He describes his competitors’ discourse as saturated with doom and casts tight control of AI as walking away from one’s own values. The dividing line therefore does not run between the companies but straight through them: research divisions asking for guardrails their own boards reject, or at least decline to erect themselves. How many signatures come from staff below executive level, the letter does not say. Everyday operations meanwhile supply concrete grounds for concern: at Hugging Face, an autonomous agent carried out more than 17,000 individual actions over a weekend to break into internal systems.
Whether the letter finds any purchase in Washington remains open. The US government, which recently weighed an oversight body modelled on the financial regulator, has said nothing about international pacing. Whether an executive’s signature counts for more than that executive’s conduct in the market will only become clear once a concrete instrument is actually on the table.
