Anthropic told US senators and White House officials that operators linked to Alibaba’s Qwen AI lab used almost 25,000 fraudulent accounts to make 28.8 million exchanges with Claude from April to June, according to people familiar with the matter.

Anthropic said it does not offer its products in China.

Alibaba had no comment.

Knowledge distillation is a technique in which a smaller AI model mimics outputs from a larger one. AI companies said unauthorized distillation can reproduce advanced capabilities more cheaply while violating service terms.

The claim echoes earlier allegations by OpenAI involving DeepSeek, and US AI labs have begun sharing information to detect similar activity.

A House bill, the Deterring American AI Model Theft Act, would prompt consideration of sanctions, including entity-list placement, against groups found using improper query-and-copy techniques.

🔗 Source: Bloomberg

Recent Anthropic developments