On 28 July 2026, over 1,200 employees from the companies building the most powerful AI systems published a joint statement called Pacing the Frontier. The signatories, including leadership from OpenAI, Anthropic, Google DeepMind, and Meta, asked the US government to support an international effort to develop tools that could deliberately pace frontier AI development. It is not a call for an immediate pause, but a request to build the technical and governance machinery that would make slowing down possible if it becomes necessary.

What happened

The statement was published at pacingthefrontier.com on 28 July 2026 and gathered signatures over the following days. By 30 July 2026 the count had passed 1,270. Notable signatories include Dario Amodei (Anthropic CEO), Jack Clark and Jared Kaplan (Anthropic cofounders), Jakub Pachocki (OpenAI chief scientist), Mark Chen (OpenAI), Shengjia Zhao and Dawn Song (Meta), Anca Dragan (Google DeepMind AI safety lead), and Chris Olah (Anthropic).

The letter’s core argument: the leading AI companies believe they could be close to automating AI research itself, which could rapidly accelerate capability development beyond our ability to understand or control the resulting systems. Each company and country is under intense competitive pressure not to unilaterally slow that acceleration, and the world currently lacks the tools to deliberately pace frontier-wide progress. The signatories ask the US government to lead an international effort to build those tools, so the option to buy time exists if emerging risks, security concerns, or oversight gaps require it.

OpenAI endorsed the statement, writing on X that “at some point in the future, AI acceleration for frontier model development may be so high that the world will need to pace the rate of AI advancement.” Anthropic’s Amodei signed personally.

Why it matters for builders

The letter does not propose specific restrictions on what you can build or deploy today. It is about preparedness: having a mechanism ready in case recursive self-improvement or automated AI research begins to outpace safety work. Anthropic made a similar argument in its Claude writes most of Anthropic’s code disclosure, calling for frontier labs and policymakers to design a slow-or-pause mechanism before it is needed.

For builders, the practical signal is that the people inside frontier labs are publicly saying the current pace may not be sustainable and that governance infrastructure is missing. If you are building on frontier models, watch for how this translates into policy. The OpenAI/Hugging Face breach that preceded the letter by a week is already cited as evidence that containment and oversight are not keeping up with capability.

Sources

  • Pacing the Frontier, statement and signatories (28 July 2026): https://www.pacingthefrontier.com/
  • Business Insider, “Over 1,100 AI workers sign letter asking US to support tools that ‘pace the frontier of automated AI development’” (29 July 2026): https://www.businessinsider.com/ai-open-letter-automated-development-2026-7
  • Fortune, “More than 1,200 AI workers are asking for Washington’s help to build an AI slowdown plan” (29 July 2026)
  • Economic Times, “Tech Leaders Urge US to Back Global AI Risk Management Initiative” (29 July 2026)

Further reading