Nvidia, Microsoft, Meta, and over 20 other companies have signed a letter urging U.S. policymakers to avoid "premature restrictions" on open-weight artificial intelligence models. They argue that such restrictions could stifle competition and drive innovation overseas, particularly as Chinese open-weight models like Moonshot AI's Kimi K3 gain prominence.

The companies emphasize that open-weight models foster competition and ensure AI benefits are widely shared, rather than concentrated. They also point out that closed models are not inherently safer, as they can be breached or misused without detection. The letter highlights the risk of concentrating advanced AI capabilities behind a few closed systems.

While companies like Open AI and Anthropic, which are developing proprietary models and preparing for IPOs, did not sign the letter, their leaders expressed support for open access. Hugging Face recently utilized a Chinese open-weight model from Z.ai to contain a cyberattack, demonstrating the utility of open models. The signatories advocate for addressing concerns about unlawful technology distillation through targeted legal frameworks, not broad restrictions on AI innovation.