Anthropic Accuses Alibaba of Massive AI Distillation Attack

Anthropic has accused the Chinese conglomerate Alibaba Group and its Qwen AI lab of orchestrating the largest known distillation attack in the AI sector. In a June 10 letter to U.S. Senators Tim Scott and Elizabeth Warren, the company alleged that Alibaba-linked operators used approximately 25,000 fraudulent accounts to generate 28.8 million exchanges with the Claude AI system between April 22 and June 5, 2026.

The campaign targeted Claude's software engineering, task planning, and agentic reasoning capabilities. Anthropic claims these distillation attacks allow competitors to train weaker models on high-performance outputs, effectively subsidizing Chinese AI development with American research and investment. This follows similar accusations made by Anthropic in February against other Chinese firms, including DeepSeek, Moonshot AI, and MiniMax.

Anthropic is now urging the U.S. government to implement financial penalties, update antitrust guidelines, and strengthen export controls on AI chips. The company argues that current export bans are ineffective against API-based siphoning. These developments coincide with broader tensions; the U.S. Department of Defense recently designated Alibaba as a Chinese military company, a move Alibaba is currently challenging in federal court.

Alibaba has not publicly responded to the distillation allegations, though its shares saw a decline in Hong Kong and U.S. markets following the reports.