Anthropic: Alibaba-Linked Operators Used 25k Accounts to Mine Claude for Qwen (runtimewire.com)

Anthropic has accused operators linked to Alibaba's Qwen AI lab of using nearly 25,000 fraudulent accounts to extract capabilities from their AI model, Claude, over a six-week period. This operation reportedly generated more than 28.8 million interactions with Claude and focused on critical functionalities like software engineering and agentic reasoning. Anthropic frames its allegations as part of a broader effort to discuss the implications of unauthorized access to AI models, arguing that these practices represent a significant national security risk and could undermine competitive integrity in the AI market. The significance of this claim lies not only in its volume but also in the potential consequences for AI development and regulatory policies. By accusing a major technology company of such actions, Anthropic is elevating the conversation around model extraction to the attention of policymakers and suggesting that current export control measures may need to be adapted to address these sophisticated distillation attacks. As the competition in AI technology intensifies, the ability to protect model outputs is crucial, making the distinction between legitimate research and exploitation increasingly important. This incident underscores the evolving tactics in the AI landscape and highlights concerns over intellectual property, market fairness, and international competitiveness.