Anthropic CEO Denies Advocating for AI Model Bans

Anthropic founder and CEO Dario Amodei issued a clear statement on Monday, refuting industry rumors that his company supports banning open-weight AI models. He emphasized that he has never advocated for such restrictions, believing open-weight models that do not possess dangerous capabilities are a public good. These models offer value to businesses, developers, and researchers without significant cost beyond computational power.

Amodei articulated long-standing concerns about AI's potential misuse by authoritarian governments, particularly for military superiority or domestic repression, identifying China as the most capable actor. He also voiced fears of AI enabling large-scale biological attacks, where the lack of guardrails and monitoring on open-weight models makes them particularly risky. He noted that once released, these models cannot be recalled.

Despite these concerns, Amodei proposed constructive actions, such as restricting China's access to advanced chips and cracking down on AI model distillation. He also expressed support for global model safety testing organizations, including China's participation, seeing potential for cooperation in preventing AI-driven biological weapons.