Beijing pushes back against US claims over AI model training.
China has warned it will respond if the United States follows through on threats to sanction Chinese artificial intelligence companies, deepening an already tense dispute over how advanced AI models are developed and trained.
The warning came Monday after China's Ministry of Commerce responded to reports that senior US officials are considering investigations into Chinese AI firms over allegations they used American frontier models to train their own systems through a technique known as model distillation.
In a statement, the Ministry of Commerce accused Washington of politicizing technology and trade while unfairly targeting Chinese companies.
"These practices lack factual basis, lack of legal support, and engage in double standards in practice, which is a typical act of artificial intelligence hegemony," a spokesperson for China's Ministry of Commerce said.
The ministry also rejected claims that Chinese AI companies had stolen US intellectual property, arguing that Chinese developers have invested heavily in foundational AI research and have achieved leading capabilities in areas including front-end coding.
The spokesperson added that innovation should not belong exclusively to any one country, saying, "Innovation is not a patent for either party."
China also claimed that model distillation is widely used across the AI industry and said many US companies have also relied on Chinese models during research and development. However, it did not identify specific firms or provide evidence.
Sanctions dispute grows
The comments follow statements from US officials signaling closer scrutiny of Chinese AI developers.
According to Bloomberg, Treasury Secretary Scott Bessent said the US would examine Chinese AI models for signs of intellectual property theft, echoing concerns previously raised by OpenAI and Anthropic that Chinese competitors may have used outputs from leading US models to build rival systems more cheaply.
Beijing dismissed those accusations and urged Washington to abandon what it described as a campaign of smearing Chinese AI firms.
The ministry also pointed to opposition within the US technology industry, saying nearly 200 American startups have urged the US government not to restrict access to Chinese open-source AI models because doing so could weaken the competitiveness of US companies.
It also said some large US multinational companies consider model distillation a standard industry practice.
Alongside its criticism, Beijing called for greater cooperation on AI governance.
The ministry urged the United States to "stop smearing and sanctioning the threat of sanctions against Chinese enterprises" and said both countries should strengthen dialogue and communication on artificial intelligence while working together to improve AI governance.
It also warned that "for any act that substantially harms China's interests, China will take all necessary measures to firmly safeguard its legitimate and lawful rights and interests."
What the dispute means
The latest exchange highlights how AI development has become another flashpoint in the broader US-China technology rivalry. Beyond competition over chips and computing power, attention is increasingly shifting to how companies train foundation models and whether techniques such as distillation cross intellectual property boundaries.
For AI companies, the outcome could influence access to open-source models, cross-border research partnerships, and future regulatory scrutiny.
At the same time, the dispute underscores a challenge facing policymakers: protecting intellectual property while preserving collaboration and innovation in a field that advances through shared research and widely adopted technical methods.
Any escalation in sanctions could make that balance even harder to achieve.
Also read: Nvidia, Microsoft, OpenAI, and Google signed an open-weight AI letter urging US policymakers to avoid broad restrictions on downloadable models.