Three open letters reflect divergent AI policy priorities on open-weight models and frontier pacing

A Microsoft-led coalition of 235 AI companies, including NVIDIA, Amazon, Y Combinator, The Linux Foundation, and OpenAI, published a letter titled *Open Weights and American AI Leadership* on July 24, 2026. The letter argues against government restrictions on open-weight AI models, asserting that reliance on closed models is not inherently safer and that open-weight models enable broader scrutiny and improvement by the research community.

Anthropic published a response titled *Our position on open-weights models* on July 27, 2026, which did not sign the Microsoft-led letter. Anthropic CEO Dario Amodei warned that authoritarian governments could exploit open-weight models to build more powerful systems and that models could be misused for cyberattacks or biological attacks. Amodei also called for a crackdown on industrial-scale distillation operations, while clarifying that Anthropic does not advocate for a ban on open-weight models.

A third letter, *Pacing the Frontier*, was published on July 28, 2026, and signed by 1,324 employees from frontier AI companies. Signatories included Jakub Pachocki (Chief Scientist at OpenAI), Ilya Sutskever (Safe Superintelligence Inc.), Dario Amodei (Anthropic), and Jack Clark (Anthropic). The letter urges the U.S. government to support an international effort to develop technical and governance tools to deliberately pace the frontier of automated AI development, citing concerns over competitive pressure and accelerated progress driven by automated research.

The letters reflect divergent priorities: the Microsoft-led coalition emphasizes open-weight models as a safeguard and competitive necessity, Anthropic focuses on mitigating risks from open-weight models and distillation, and the employee-led letter calls for international governance to manage rapid AI advancement.