Ahead of the upcoming Trump-Xi summit, where artificial intelligence is expected to be a key topic, China’s state-run People’s Daily published a commentary today pushing back against recent U.S. accusations that Chinese AI firms are engaged in “industrial-scale” distillation of American AI models.
The article, written under the byline “Zhong Sheng (钟声)” — a pseudonym for “China’s Voice,” frequently used by the paper for authoritative statements reflecting the Chinese government’s position — claims that a joint warning issued by the U.S. National Security Agency, the Cybersecurity and Infrastructure Security Agency, and the FBI has no legal basis. It argues that distillation is a standard practice used by AI companies worldwide, including American firms, and characterizes the U.S. crackdown as disguised protectionism rather than a genuine security concern. The commentary accuses Washington of applying a double standard — treating the same technique as routine when used by U.S. firms but as an “attack” when used by Chinese firms — and warns that Beijing will retaliate if its companies’ legitimate interests are harmed.
The U.S. agencies had alleged that Chinese firms including DeepSeek, Moonshot AI, Alibaba, MiniMax, and StepFun used variants of American AI models to train their own products, likely with the Chinese government’s knowledge, and described the activity as aggressive and malicious.
Distillation refers to using outputs from large, costly AI models to train smaller ones more cheaply. The dispute comes as the two countries prepare for a second Trump-Xi summit in Washington on the 24th, with AI expected to be a major topic; Reuters has reported the two sides plan a separate dialogue on AI safety risks in mid-September.
Source: Central News Agency (Taiwan), September 16, 2026
https://www.cna.com.tw/news/acn/202609160111.aspx