Anthropic Accuses Alibaba of Largest Distillation Attack on AI Models

Anthropic has accused Alibaba of conducting the largest known distillation attack on its Claude AI models, using 25,000 fraudulent accounts to execute 28.8 million exchanges. This attack, described in a letter to U.S. Senators, involved illicitly extracting capabilities from Claude to enhance Alibaba's own AI models. The operation is seen as a systematic effort to harvest U.S. AI capabilities without incurring the associated research and development costs. Anthropic's head of policy, Sarah Heck, has called for legislative measures to curb such attacks, including limiting Chinese access to advanced U.S. computing resources. The incident follows recent U.S. export controls on Anthropic's latest AI model, Fable 5, due to national security concerns. Alibaba has also faced scrutiny from the Pentagon, being added to a blacklist linked to the Chinese military. As of the latest reports, Alibaba's stock has dropped over 4% following these allegations.

Key Points: • Alibaba allegedly conducted the largest distillation attack on Anthropic's Claude AI models. • The attack involved 25,000 fake accounts and 28.8 million exchanges with Claude. • Anthropic is urging U.S. lawmakers to implement stricter regulations against such attacks.