Anthropic blocks Chinese users from AI models amid distillation attack fears

US artificial intelligence company Anthropic has embedded advanced code in its AI models to prevent access by Chinese users, citing national security concerns. The company has identified and suspended accounts suspected of being operated by Chinese entities and is actively working to disrupt proxy networks used to bypass restrictions. The move follows allegations that Chinese tech giant Alibaba conducted a large-scale 'distillation attack' aimed at replicating Anthropic's advanced AI capabilities at a significantly lower cost. Such distillation techniques allow rivals to clone a model's performance by training on its outputs, rather than developing the technology independently.