Twenty-four organizations and companies, including Meta, Microsoft, Nvidia, IBM, and Mistral, are urging U.S. policymakers to protect open AI models. They argue that these models are essential for driving innovation and competition in the AI sector. In a joint letter led by Nvidia CEO Jensen Huang, the coalition warned against broad restrictions on downloadable AI models. Targeted measures, they said, would be a more effective way to address security concerns and copyright risks.

The group called on lawmakers to expand access to computing resources for startups and researchers, as well as to fund shared training datasets and evaluation frameworks. While the initiative is partly a response to recent Chinese AI models, it may also reflect the interests of major infrastructure and chip companies that want to encourage a wider range of usable models in order to increase revenue from computing capacity and related services.