What happened
Liu Liehong, head of the National Data Administration, said at the 2026 China International Big Data Industry Expo that business models should be explored around major scenarios in the national economy, including token value-added subscriptions and pay-by-performance arrangements.
He explained that data annotation is like labeling raw data, and the annotated data can then be used to train artificial intelligence. A new batch of data annotation pilot programs will be set up in 32 cities, with a focus on fields such as industrial manufacturing, healthcare and transportation, to address bottlenecks in the industry.
Liu also said multiple departments will strengthen coordination on digital talent development, pushing for an integrated training mechanism linking undergraduate, master's and doctoral programs for data-element-related disciplines and improving the overall discipline layout.
Why it matters
The remarks signal growing policy attention to data annotation as a foundation for AI development. Exploring new commercial models could make annotated data services more sustainable and encourage wider participation.
Expanding pilots to 32 cities and several industries suggests an effort to standardize and scale data annotation across real economic sectors, while the proposed education pathway points to longer-term planning for talent supply.
Key facts
Liu Liehong made the remarks at the 2026 China International Big Data Industry Expo.
He proposed exploring token value-added subscriptions and pay-by-performance business models around major national economic scenarios.
Data annotation means labeling raw data so the labeled data can be used to train AI.
The National Data Administration will launch a new batch of data annotation pilots in 32 cities.
Pilot areas include industrial manufacturing, healthcare and transportation.
Multiple departments will work to establish a bachelor's-to-doctoral training mechanism for data-element-related disciplines.
What to watch next
How the 32-city pilots translate into practical business models for data annotation services.
Whether cross-department coordination leads to new education programs and career pathways in data-related fields.