US Accuses Six Chinese AI Firms of Misusing American AI Models

US security agencies have accused six Chinese artificial intelligence companies of carrying out large-scale efforts to extract capabilities from American frontier AI models. The companies named include DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI, according to reports published this week.

The allegations centre on a technique known as model distillation, in which information generated by a more capable AI system can be used to help develop or improve another model. US authorities allege that the Chinese firms obtained large volumes of responses from American AI systems as part of efforts to accelerate their own research and development.

The accusations come amid intensifying competition between the United States and China over artificial intelligence. Chinese developers have rapidly expanded their presence in the global AI market, with open-weight models from companies such as DeepSeek, Alibaba and Z.AI gaining significant adoption internationally.

The dispute also highlights growing concerns about how AI companies protect their models and intellectual property. US technology companies have previously accused Chinese competitors of improperly obtaining model capabilities, while Chinese firms have continued to advance rapidly and narrow the performance gap with leading American systems.

The latest allegations are likely to add pressure to an already tense technology relationship between Washington and Beijing. The Chinese companies have not been established guilty by these allegations, and further evidence and responses from the companies involved will be important as governments consider how to protect advanced AI technology while maintaining international competition and innovation.