Anthropic accuses China's Alibaba of largest AI 'brain-theft' operation ever recorded

Anthropic, in its threat intelligence report, accused China's Alibaba of largest "illicit" AI distillation campaign ever detected. The campaign targeted chain-of-thought reasoning transcripts from Opus 4.6 and 4.7, extracting Claude's reasoning at a peak rate of nearly 3 million exchanges per day. It used more than 3,500 fraudulent accounts throughout the campaign to train Alibaba's AI models.