Nvidia and Meta join a policy letter advocating open AI models, potentially eroding proprietary model advantages held by OpenAI and Anthropic.
Nvidia (NVDA) and Meta (META) have signed a policy letter urging U.S. policymakers to avoid broad restrictions on open-weight AI models. The letter, published July 24, aims to allow models whose weights can be downloaded, modified, and run privately. OpenAI and Alphabet (GOOGL) later joined the effort, per reports on July 27.
The move contrasts with Anthropic, which remains absent from the letter. Open-weight models enable developers to host and adapt AI locally, reducing reliance on proprietary APIs. Nvidia stands to benefit as increased model development and deployment could drive demand for its AI training and inference hardware.
Meta’s support aligns with its strategy to integrate open models into products and corporate systems, competing with OpenAI and Anthropic. While closed models may retain advantages in performance and support, open alternatives could pressure pricing and market share for proprietary AI services.