A coalition urged US policymakers not to broadly restrict open-weight AI models, arguing that limits could undermine competition and push development overseas. (Published November 14.)

On July 24, 2026, 25 technology companies—including Microsoft, Nvidia, Meta, IBM, Palantir, Mistral and Hugging Face—published “Open Weights and American AI Leadership,” urging US policymakers to avoid premature restrictions on open-weight AI. The letter says such models can be downloaded, inspected, modified and run on a user’s own infrastructure, widening access for startups and researchers. Signatories argue broad limits could “stifle competition or drive innovation overseas,” while concentrating advanced capabilities in a few closed systems; they say misuse and unlawful distillation should be handled with targeted legal and commercial measures. The stakes extend beyond model makers: cloud and chip providers, developers and businesses deploying AI all benefit when models are adaptable and providers compete on cost and services. The companies also called for expanded access to compute and shared datasets, tools and evaluation frameworks.