Our position on open-weights models

Anthropic CEO Dario Amodei said the company does not support banning open-weights AI models.

He said open-weights systems can be useful public goods, but may pose higher misuse risks than closed models.

Amodei argued that US limits on Chinese open-weights models would not address his core concerns.

Instead, he called for tighter chip export controls to China, stronger anti-smuggling enforcement, and crackdowns on large-scale model distillation.