China tells Washington to drop the AI theft accusations
Beijing calls US claims that six Chinese AI labs distilled American frontier models "baseless" — days before a Trump-Xi meeting.
Beijing wants the accusations to stop. A day after the FBI, NSA and CISA issued a joint advisory accusing six Chinese AI companies — DeepSeek, Alibaba, Moonshot AI, MiniMax, StepFun and Z.AI — of "industrial-scale" theft of US trade secrets, China's foreign ministry pushed back, with spokesperson Mao Ning calling the country's AI progress the product of "independent scientific research" and urging Washington to drop what it calls baseless claims.
The technical charge is distillation: training a weaker model on the outputs of a stronger one to "extract restricted proprietary functionalities" from US frontier systems. The advisory says it has been going on since 2024, against Anthropic's Claude, OpenAI's ChatGPT, Google's Gemini and xAI's Grok. Anthropic's own numbers are the sharpest detail — roughly 24,000 fraudulent accounts generating over 16 million exchanges with Claude. Alibaba's Qwen family is named specifically.
This isn't a new fight. In July, China's Ministry of Commerce made the mirror-image claim: that US labs distil Chinese models. What's changed is timing — the advisory lands weeks before a scheduled Trump-Xi meeting, with Congress floating sanctions legislation.
Why it matters: distillation is nearly impossible to prove from the outside, which makes it the perfect pretext for export controls that were coming anyway.
Sources
- US claims Chinese AI firms are carrying out 'industrial-scale' theft of trade secrets cnn.com
- Six Chinese AI Companies Accused by US Intelligence of Tech Theft via Model Distillation parameter.io
Written by an AI pipeline from the sources above. How it works.
Feed, daily deep-dive and bytes — readable offline, with push alerts for the topics you follow.