Amodei Denies Anthropic Supports Open-Weight AI Ban, Warns of China Risks

Anthropic CEO Dario Amodei pushed back Monday on suggestions that his company supports banning open-weight AI models, stating clearly that Anthropic has never advocated for such bans. His comments followed an open letter from Nvidia CEO Jensen Huang and other companies, including Hugging Face, Meta, Microsoft, and Mistral, urging policymakers to avoid broad restrictions on open-weight models amid concerns about Chinese AI firms allegedly using distillation techniques to replicate American models.

Amodei said he views open-weight models without dangerous capabilities as a public good, distinct from broader concerns about authoritarian governments, particularly China, potentially using AI to gain military advantages or enable repression. He said his primary concerns center on scenarios involving biological or cybersecurity threats, arguing open-weight models are harder to monitor or restrict once released.

Rather than supporting a ban, Amodei called for continued restrictions on advanced chip exports to China and stronger action against unauthorized distillation. He also expressed support for developing a global AI safety testing framework, one that would ideally include China’s participation, suggesting shared interest in preventing AI-enabled biological weapons could make such cooperation possible despite broader geopolitical tensions.