Anthropic accuses Alibaba of illicitly extracting its AI tech via massive distillation attack.

Anthropic has accused Alibaba of conducting the largest known distillation attack on its AI models, using 25,000 fraudulent accounts to make 28.8 million exchanges between April and June. Distillation involves creating a smaller AI model by extracting knowledge from a more advanced one, which Anthropic claims Alibaba did illicitly to gain its AI capabilities. This accusation comes amid broader concerns about industrial-scale AI model theft and follows government warnings and export control directives related to AI technology access. Anthropic is working with U.S. policymakers to address these threats and maintain American leadership in AI innovation.