Anthropic CEO Dario Amodei clarifies position on open-weights models, opposing bans while supporting chip controls and mandatory safety testing

Dario Amodei published an official Anthropic position statement clarifying that the company has never advocated for banning open-weights models, pushing back against accusations of protectionism amid US policy debates about Chinese AI models. Amodei outlines two primary national security concerns — authoritarian AI superiority and misuse for cyberattacks or bioweapons — and argues that blanket bans address neither, while endorsing chip export controls, crackdowns on distillation operations, and mandatory safety testing for all sufficiently capable models regardless of origin or openness. The statement engages directly with an industry open letter on open-weights models and references Anthropic's own research on modular training strategies as a potential safety mitigation path.