"Slow Down AI Development" OpenAI and Anthropic Developers Urge U.S. Government

1,224 employees from OpenAI, Anthropic, and other frontier AI firms signed “Pacing the Frontier,” urging the U.S. to support international technical and governance tools to slow or coordinate advanced AI development.

The statement warns that AI research automation and recursive self-improvement may arrive soon, while current safety and oversight methods are not enough to keep up.

Supporters, including Anthropic CEO Dario Amodei, say the goal is to pace AI so society can adapt, not to reopen the open-versus-closed model debate.

The call comes as strong open-weight models from China and broader industry caution are intensifying debate over how to manage frontier AI safely.