The United States has accused six Chinese AI firms—DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI—of aggressively copying leading US AI models since late 2024. The NSA, CISA, and FBI allege these firms could shave billions off their development costs by waging industrial-scale attacks on US models.

According to the agencies, these Chinese firms likely acted with Chinese government awareness when extracting capabilities from US models, including variants of Claude, GPT, Gemini, and Grok. This has led to significantly shorter AI development timelines and reduced financial expenditures for Chinese companies.

To combat this, American AI firms must work with the government and US allies. Agencies recommend mitigations that would make it harder for Chinese firms to steal from US models, such as improving detection of sophisticated campaigns using tens of thousands of accounts and routing distillation requests through multiple pathways.

The agencies state that all American AI firms must take action to end the alleged theft, which threatens the US lead in the AI race. This will require coordinated action across the AI ecosystem to combat the aggressive, malicious, and targeted distillation activities.