Anthropic Accuses Alibaba of ‘Illicitly’ Accessing Its Claude AI Models in Largest Known Distillation Attack

Anthropic alleges that Alibaba and its AI research division conducted an industrial-scale adversarial distillation campaign from April 22 to June 5, 2026, using approximately 25,000 fraudulent accounts and 28.8 million exchanges to illicitly harvest Claude’s advanced capabilities (including software engineering and agentic reasoning); Anthropic warns this enables replication of frontier U.S. AI without R&D costs, poses safety and national security risks, and has spurred potential legislative and sanction actions.