Anthropic Claims Alibaba Ran 'Brazen' Campaign to Access Its Claude AI Model

- Anthropic (maker of Claude) and Alibaba (China's giant e-commerce/cloud firm) are direct rivals in the global AI race. Alibaba has its own AI models, the Qwen family. - Anthropic has accused Alibaba of systematically trying to steal its proprietary AI technology — specifically by creating fake customer accounts to access Claude's internal "weights" (the mathematical parameters that define how the model behaves, its core intellectual property). - A model's weights are its most valuable trade secret; copying them lets a competitor effectively clone the model without doing the expensive research and training themselves. - This is part of a broader pattern: US AI companies increasingly accuse Chinese firms of IP theft, while Chinese firms say the US is trying to block their technological rise. The US government has also restricted exports of advanced AI chips to China. - The accusation is notable because Anthropic is going public with a legal threat rather than quietly settling, signaling escalating tension in the US-China AI rivalry.