OpenAI, Anthropic staff urge US to help pace frontier AI

More than 1,000 employees at the biggest AI labs put their names on a public letter called Pacing the Frontier this week, and the interesting part is who they are and what they are actually asking for. As The Verge reported, the signatories include Anthropic CEO Dario Amodei, OpenAI chief scientist Jakub Pachocki, Meta chief scientist Shengjia Zhao, and Google's head of AI safety Anca Dragan, plus co-founders and senior staff from the same handful of frontier labs. Anthropic and OpenAI backed it as companies within hours.

The single ask is narrower than a lot of the coverage implied, and worth quoting. The statement requests that the US government "support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development." That is a request to build the brake, not to pull it. According to the reporting, the signatories are explicit that they are not calling for a pause or slowdown now, only that the option to pace exists later.

The framing matters because these are the same executives shipping the models. What the letter is really saying is that no individual lab or country can slow down unilaterally without conceding competitive ground, so the tooling to coordinate has to come from government. That is a candid admission about the shape of the race, and it puts unusual political air cover behind treating automated, self-improving AI as its own policy category rather than folding it into general AI regulation.

The honest caveat is that a statement is a statement. The retrieved reporting does not describe any concrete mechanism, capability threshold, or enforcement body attached to it, and it does not say what happens if Beijing, Brussels, or the UK AI Safety Institute do not play. Skeptics will note that governance tools designed with heavy input from Anthropic and OpenAI have obvious incumbent-friendly failure modes.

If you are watching the space, the useful move is to treat this as pre-work for whatever the next US Congress or administration does on frontier AI. The statement itself is short; the political fight over what "pace" means in practice is where the real stakes live.