Anthropic Accuses Chinese Firms Of Ai Distillation Attacks

On February 24, 2026, Anthropic publicly accused three Chinese AI firms—DeepSeek, Moonshot, and MiniMax—of conducting extensive distillation attacks against its Claude chatbot, claiming that these companies engaged in campaigns to illicitly extract Claude's capabilities through over 16 million interactions via approximately 24,000 fraudulent accounts. Anthropic stated, "We’ve identified industrial-scale distillation attacks on our model," emphasizing the severity of the situation as a national security threat that necessitates urgent action from both government and industry. The company linked these activities to the firms with high confidence, citing evidence such as IP address correlations and metadata requests, which were corroborated by other industry participants observing similar patterns of behavior. This revelation follows earlier accusations from OpenAI, which reported similar issues with rival firms distilling its models and subsequently banning accounts suspected of misuse.