China NDA Plans Embodied AI Data Standards After Industry Ask

China's National Data Administration says it will develop standards for embodied AI training data and guide local authorities on the work, ten days after seven companies asked the same agency for exactly that. The Next Web reported that Liu Liehong, who heads the NDA, chaired a Sept. 10 symposium with research institutes, technology companies and humanoid robotics organizations; the agency posted its account on Sept. 13.

The seven companies "requested public data infrastructure for embodied AI and common data standards," according to the agency, which said "the industry is becoming more data-driven."

The gap those standards would sit on is wide. Embodied AI foundation models need approximately 10 million hours of real training data, according to the China Academy of Information and Communications Technology. Only 100,000 to 1 million hours of high-quality data currently exist globally.

China's physical footprint is already running ahead of that number. The country operates more than 70 embodied AI training grounds across over half its provincial regions, with 46 additional facilities planned or under construction. About 86% target industrial manufacturing applications, concentrated in three clusters: the Yangtze River Delta, the Beijing-Tianjin-Hebei region and the Pearl River Delta.

Europe has the rules but not the collection layer. The Data Act took effect September 12, 2025, governing data access from connected products, and planned data labs from the November 2025 Data Union Strategy remain unestablished.

No draft standard, deadline or public consultation was disclosed. In February the Ministry of Industry and Information Technology's technical committee released a six-part Humanoid Robot and Embodied Intelligence Standard System (2026 Edition); the NDA's move slots data-specific rules alongside that framework. It lands in a busy fortnight of Chinese AI policy activity on our tracker.