Why is Chinese state media telling people to stop using English AI terms?

China's state-run People's Daily has urged the public to replace English AI terms with standardized Chinese equivalents, aiming to strengthen the country's "discourse power" in global technology discussions. As artificial intelligence spreads across China, English words such as "agent" and "LLM" for large language models have become commonplace, the commentary highlighted.

The article warned that reliance on foreign terminology could create "cognitive dependency" and diminish China's influence in global technology governance. This push comes after China's official adoption of "ciyuan" as the standard translation for "token" in March at the China Development Forum 2026, where "ci" means "word" and "yuan" is associated with fundamental elements.

The commentary argued that English AI terms are often unclear to ordinary Chinese people, unlike earlier technology names that were localized, such as "dian nao" for computer. Untranslated terms risk widening the digital divide and eroding the Chinese language system's integrity. Building up China's "discourse power" has been a priority, involving shaping narratives through stronger external propaganda and domestic frameworks.

While some experts find the push for local translations as a symbolic gesture, the People's Daily emphasized a dual-track approach: retaining English terms in international exchanges while promoting standard Chinese translations in domestic contexts like education and policy. This approach seeks to strike a balance between efficient global communication and cultivating a strong native language foundation among the Chinese people.