China accuses US AI firms of training on Chinese models

China on Monday accused American artificial intelligence companies of using Chinese models to train their own systems, sharply escalating a technological dispute days after U.S. officials threatened sanctions against Chinese firms for alleged intellectual property theft. The Chinese Ministry of Commerce issued a statement rejecting the U.S. allegations as lacking factual and legal basis and condemned what it called acts of "artificial intelligence hegemonism."

The ministry asserted that many U.S. AI enterprises have themselves used a technique called distillation, which involves training an AI model to mimic the output of a larger, more capable one, on Chinese models during their development. The statement warned that China would take all necessary measures to safeguard its legitimate rights and interests against any actions causing substantive harm.

This diplomatic exchange follows accusations leveled by U.S. officials last week against the Chinese AI startup Moonshot AI. A senior White House official alleged that Moonshot covertly copied Anthropic's most advanced model, known as Claude, to build its own recently released Kimi K3 model. U.S. Treasury Secretary Scott Bessent...