Microsoft, Nvidia and peers urge U.S. support for open-weight AI

Microsoft, Nvidia, Meta, Hugging Face and more than 20 other technology organizations published an open letter urging U.S. policymakers to back open-source and open-weight AI systems instead of favoring closed proprietary models or imposing broad access restrictions. The companies said open models would widen access for startups, universities and research labs, reduce vendor lock-in, and let enterprises and governments run AI on their own infrastructure while keeping sensitive data behind their own firewalls.

The push lands amid an active Washington debate over AI policy, including possible restrictions on some Chinese open models and the White House Gold Eagle AI cybersecurity clearinghouse for AI-driven vulnerability discovery and remediation. Supporters argued that broader access to model weights would strengthen innovation and give defenders capabilities closer to those available to attackers, while critics warned that open models can also lower the barrier for cybercrime, enable removal of safety guardrails, and facilitate abuse such as deepfakes and CSAM-related content.