Anthropic CEO Dario Amodei responded to industry speculation that his company supports banning open-weight Chinese AI models, or open-weight models generally.

In a blog post, Amodei explicitly stated, “Anthropic has never advocated for a ban on open-weights models,” emphasizing his own wording.

His clarification followed an open letter from Nvidia CEO Jensen Huang and other companies including Hugging Face, Meta, Microsoft, Mistral, and Nvidia, which urged policymakers to avoid “premature restrictions” on open-weight AI models.

While the letter did not mention China, the industry debate has centered on allegations of Chinese AI labs using techniques like distillation to learn from U.S. models.

Amodei argued that open-weight models without dangerous capabilities are a public good, providing value to businesses, developers, and researchers.

His fears focus on authoritarian governments building more powerful models to achieve “permanent military superiority” or repress their own people, citing the Chinese Communist Party (CCP) as the most capable concern.

He also warned that AI could enable biological attacks and that open-weight models are harder to guardrail or monitor, noting a UK AI Security Institute report that once released, these models cannot be withdrawn.

This counters open-source advocates who argue that broad access helps defenders. Amodei proposed restricting China’s access to powerful chips (a current U.S.

policy) and cracking down on distillation.

He also supported creating a global model safety testing organization, where China would need to participate, citing limited cooperation on preventing AI biological weapons as possible because it aligns with China’s interests.