Several major AI companies have urged US policymakers not to introduce broad restrictions on open-weight AI models while Washington considers how to respond to alleged intellectual-property theft by Chinese AI labs. The letter argues that any action should distinguish unlawful extraction from closed models from legitimate uses of model distillation, warning that sweeping rules could hinder innovation, transparency and cyber defence.

The signatories, including Hugging Face, Meta, Microsoft, Mistral and Nvidia, say open models can help defenders test and address AI-enabled threats. They cite Hugging Face’s recent use of Chinese firm Z.ai’s open-weight GLM 5.2 after commercial frontier models’ safeguards prevented it from developing defensive tools; the debate follows reports that the Trump administration may ban Chinese open-weight models and sanction Chinese AI companies.