China’s National Data Administration announced plans to develop standards and guide regional authorities on embodied AI data, responding closely after companies formally requested a unified data framework and infrastructure to support this emerging sector.
- China to create embodied AI data standards after industry request
- Over 70 AI training grounds operational, with more planned
- Europe regulates data use but lacks embodied AI data initiatives
What happened
China’s National Data Administration revealed it will develop formal standards for embodied AI data and assist local authorities in overseeing this work. This announcement came just ten days after seven companies petitioned the agency to establish public data infrastructure and common standards for AI training datasets.
The announcement followed a high-level meeting organized on September 10 that involved research institutes, major technology firms, and humanoid robotics organizations. The regulator emphasized a shift towards a data-driven AI industry and expressed support for companies investing more in data resource development.
Why it matters
Embodied AI technologies require extensive real-world training data, estimated at about 10 million hours to build foundational models, while current global data availability is far less. China’s move to set standards and expand data collection sites addresses a significant bottleneck in AI advancement and could boost its position in global robotics.
More than 70 embodied AI training grounds already operate across China’s provinces, with an additional 46 planned, mostly focusing on industrial manufacturing in clusters like the Yangtze River Delta and the Beijing-Tianjin-Hebei region. This contrasts sharply with Europe, which regulates data access via the Data Act but lacks equivalent infrastructure or data production programs.
What to watch next
Progress on implementing China’s standards and the construction of new embodied AI training grounds will be key indicators of the country’s capacity to accelerate data-driven robotics innovation. Observers should monitor how local authorities enact guidance and how effectively companies increase investment in data resources.
Meanwhile, Europe’s Data Union Strategy aims to scale AI data access but has yet to operationalize promised data labs and infrastructure. The ability of European and US firms like NEURA Robotics to expand their training gym networks may influence regional competitiveness in embodied AI development alongside China’s advancing ecosystem.