Anthropic Responds to Open Source Controversy: Opposes Blanket Bans on Open Models, Advocates for Restrictions on Chips and Model Distillation

According to Odaily, Anthropic CEO Dario Amodei has addressed the controversy surrounding open-weight models, stating that the company has never advocated for a blanket ban on open models. Previously, companies including OpenAI, Google, and SpaceX joined an open-source coalition, leaving Anthropic as the only major frontier AI company not to sign on.

Amodei acknowledged that open-weight models can reduce costs, promote competition, and support customer self-deployment. He noted that open models without dangerous capabilities can be viewed as "public goods."

However, he argued that open models are not necessarily safer, and defenders do not necessarily gain more advantages than attackers. Once model weights are made public, safety restrictions can be removed and cannot be revoked.

Anthropic has proposed three alternative measures: restricting the flow of advanced chips and manufacturing equipment to China, cracking down on industrial-scale model distillation, and requiring all models—both open and closed-source—that reach a certain capability threshold to undergo cybersecurity, biological risk, and alignment testing.