Anthropic CEO Amodei clarifies stance on open-weight AI, voices concerns over China

Anthropic CEO Dario Amodei clarified that his company does not support banning open-weight AI models, contrary to some industry speculation. He stated that open-weight models without dangerous capabilities are a public good. Amodei's comments followed an open letter from Nvidia CEO Jensen Huang and other AI companies, including Meta and Microsoft, urging policymakers against premature restrictions on open-weight models. While that letter did not specifically mention China, the debate has been fueled by concerns over intellectual property theft by Chinese AI labs. Despite his support for open-weight models, Amodei expressed significant fears about authoritarian governments, particularly China, developing powerful AI for military superiority or to suppress their populations. He also warned of the potential for AI to enable biological attacks, noting that open-weight models are harder to control in such scenarios. To counter these threats, Amodei suggested restricting China's access to advanced chips and cracking down on AI distillation techniques. He also voiced support for global AI model safety testing initiatives, suggesting that even the Chinese Communist Party might cooperate on preventing AI-related biological weapons, as it aligns with their interests.