Anthropic has finally responded to the open-source controversy: Opposes banning, but calls for Chipper and Distillation

According to DuckDuckGo Beating monitoring, Anthropic CEO Dario Amodei responded to the open weight model controversy, stating that the company has never advocated for a complete ban. After OpenAI, Google, and SpaceX successively joined the open-source declaration, Anthropic remains the only major leading model company that has not signed.

Amodei acknowledged that open weight models can reduce costs, strengthen competition, and allow customers to deploy them on their own. He stated that open models without dangerous capabilities are a "public good."

However, he does not agree that open models are always safer, nor does he agree that defenders necessarily benefit more than attackers. Once weights are released, security constraints can be removed, and models cannot be retracted.

Anthropic has proposed three alternative solutions: restricting the flow of advanced chips and manufacturing equipment to China; combating industrial-scale model distillation; requiring all sufficiently powerful models to undergo network attack, biosecurity, and alignment tests, regardless of whether they are open or closed source.