Several major technology companies, including Hugging Face, Meta, Microsoft, Mistral, and Nvidia, have signed an open letter urging U.S. policymakers to avoid imposing sweeping restrictions on open-weight AI models. The letter arrives amid reports that the Trump administration is considering banning Chinese open-weight models following allegations that firms like Moonshot AI improperly used Anthropic’s Fable model to develop its Kimi K3 system.
The letter argues that distillation, a common technique for improving and evaluating AI models, should not be conflated with unlawful misappropriation, and that legitimate concerns should be addressed through targeted legal frameworks rather than broad bans. Replit CEO Amjad Masad, a signatory, said restricting Chinese open models would effectively amount to banning open models altogether, noting the interconnected nature of the open-source AI ecosystem.
The letter also defends open-weight models against claims that they pose heightened cybersecurity risks, arguing that defenders benefit from having access to capable models for detecting and responding to threats. This point follows Hugging Face’s disclosure that closed AI models refused to help it defend against a recent AI-driven attack, prompting the company to rely on an open-weight model from Chinese firm Z.ai instead.
Notably absent from the letter were closed-model developers OpenAI, Anthropic, and Google DeepMind.
