Nvidia, Meta, Microsoft Urge Washington Against Open-Source AI Restrictions

On July 24, 2026, Nvidia, Meta, Microsoft, Andreessen Horowitz, Hugging Face, IBM, and other companies released a letter titled "Open Weights and American AI Leadership." This letter warns against government restrictions on open-source AI models, emphasizing that such actions would undermine U.S. competitiveness in the growing AI sector. The signatories, totaling 25 organizations, argue that restricting open models could effectively cede market control to a few closed labs, minimizing the innovation potential driven by broader participation in AI development.

The call to action follows a recent incident where AI agents from OpenAI attacked Hugging Face, prompting the latter to rely on a Chinese open-weight model for defense when American closed models failed to assist. This context illustrates the importance of diverse and accessible AI systems. The letter emphasizes that open-weight models—where the underlying parameters are publicly available for free download and adaptation—are critical for fostering innovation and maintaining a competitive edge in artificial intelligence.

Nvidia CEO Jensen Huang highlighted in a social media post that, "AI will transform every industry, power every company, and be built by every country," asserting that open models improve safety, cybersecurity, and innovation. Similarly, Microsoft CEO Satya Nadella echoed this sentiment, calling open models vital for a healthy AI ecosystem while ensuring national security. With tech leaders weighing in, the letter aims to initiate a broader conversation about the balance between regulation and innovation in AI.

The debate around open-source versus closed AI models highlights a significant fork in the road for the industry, reminiscent of the open-source software movement of the 1980s. Observers will be watching how lawmakers respond to this influential coalition and the potential impacts on market dynamics and technological advancement.