A heavyweight coalition of AI players is drawing a line in the sand. Hugging Face, Meta, Microsoft, Mistral, and Nvidia have signed an open letter urging policymakers to steer clear of “premature restrictions” on open-weight AI models. The timing isn’t subtle. The missive lands right as the Trump administration weighs a potential ban on Chinese open models like DeepSeek, following accusations that labs including Moonshot AI distilled proprietary knowledge from Anthropic’s models to train their own highly capable systems.
The letter doesn’t name China directly, but the target is obvious. It argues that legitimate techniques like distillation—using one model’s outputs to train another—shouldn’t be conflated with outright IP theft. “Distillation… is a widely used technique for model improvement, evaluation, and validation,” the signatories write. Amjad Masad, CEO of Replit, put it bluntly to TechCrunch: “I think banning Chinese open models is as good as banning open models in general.” He pointed out that Thinking Machines Lab’s new Inkling model was trained with help from Moonshot’s Kimi 2.5. The ecosystem is deeply intertwined, and the signatories argue a ban would set a dangerous precedent.
The letter also pushes back hard on the security argument. It contends that locking down powerful models actually hurts defenders. In a stark, real-world example, Hugging Face revealed it couldn’t defend itself against a recent attack using commercial frontier models because their guardrails blocked legitimate defense efforts. The company had to pivot to Z.ai’s open-weight GLM 5.2 model to protect its own infrastructure. The subtext is clear: closed models left them defenseless.
The schism in the industry is now public. Noticeably absent from the letter are the major closed-source players: OpenAI, Anthropic, Google DeepMind, and SpaceX. Their business models are directly threatened by the proliferation of cheap, capable open models. The letter’s signatories have their own economic incentives, of course—Nvidia and Microsoft Azure profit when models become commoditized and developers buy more GPUs and cloud capacity. But the core message isn’t just about money. It’s a warning that heavy-handed restrictions would stifle competition and push the center of gravity for AI innovation overseas.
