China's Alibaba bans Anthropic AI for employees after 'distillation attack' accusation

Alibaba added Anthropic's Claude Code to its internal high-risk software list, according to a report by TechCrunch. The action followed Anthropic's public statement that Alibaba had attempted to extract Anthropic's proprietary model information through repeated queries, a technique known as model distillation. Anthropic did not identify Alibaba by name in its initial disclosure but confirmed the company had engaged in the practice. Alibaba has not publicly commented on the classification or the distillation allegation.

Next story