Policy & RegulationAnalysis
China Unveils Measures to Expand Data Industry as Sector Tops 6 Trillion Yuan
New policy initiatives target AI training pilots across 32 cities, state-firm dataset releases, and token-based commercial models.
Share
Photo by Burak The Weekender on Pexels
The Brief
China has rolled out a package of initiatives to accelerate the development of its data industry, according to the National Data Administration. The measures include establishing data-labeling pilot programs in 32 cities to support artificial intelligence training, opening access to over 6 billion pieces of high-value data held by 14 state-owned enterprises, and exploring token-based commercial models. The announcements coincided with a report released at the 2026 China International Big Data Industry Expo showing China's data sector has exceeded 6 trillion yuan, with AI-linked products and services generating more than 70 percent of total output.
Why it matters
Data has evolved from a passive informational asset into a core production factor driving China's artificial intelligence ecosystem. With AI-related offerings generating over 70 percent of an industry valued above 6 trillion yuan, government-backed efforts to release state enterprise data and organize municipal labeling hubs directly target the scarcity of high-quality training corpuses, a major constraint on machine learning development.
China context
Under the policy framework of developing 'new productive forces' and building a 'Digital China,' the National Data Administration coordinates data infrastructure and market development nationwide. By mobilizing central state-owned enterprises to supply underlying operational data and organizing city-level industrial pilots, Beijing is deploying administrative coordination to construct an integrated ecosystem linking big data reserves to artificial intelligence applications.
Editor's View
EDITOR'S VIEW — Analysis and inference, not factual reporting.
The policy package demonstrates that Chinese economic planners now view the data industry almost entirely through the lens of artificial intelligence enablement. Rather than prioritizing data management purely from a compliance or security perspective, the National Data Administration is actively testing market structures—such as token-based commercial models—to monetize and standardize data flow. Success will depend heavily on whether private AI developers can obtain meaningful, cost-effective access to state-owned enterprise data catalogs.
What to watch
- Publication of the official list of the 32 designated data-labeling trial cities and their operational guidelines
- Release of specific data catalogs, licensing terms, and access criteria by the 14 participating state-owned enterprises
- Curricular and enrollment rollouts for newly approved data-factor disciplines across Chinese universities
Key Takeaways
- 1The National Data Administration launched data-labeling trials across 32 cities to supply training data for AI models.
- 2Fourteen state-owned enterprises will make over 6 billion entries of high-value data available for operational use cases.
- 3Regulators are exploring token-based commercial models and setting up dedicated academic disciplines for the data economy.
- 4China's data industry has surpassed 6 trillion yuan, with AI-related products comprising over 70 percent of total output across more than 480,000 companies.
China has introduced a comprehensive set of policy measures to accelerate the growth of its data industry, tying the sector closely to artificial intelligence training and industrial digitalization, according to announcements from the National Data Administration reported by state media.
Under the plan, authorities will launch data-labeling trial programs across 32 cities to create structured data resources tailored for AI model training. In addition, 14 centrally owned and state-owned enterprises will open access to more than 6 billion items of high-value data, aiming to stimulate data utilization in practical industrial and commercial scenarios.
Regulators also revealed plans to explore token-based commercial models adapted to industrial needs, facilitating digital and intelligent transformation across traditional sectors. Token-based metrics, commonly utilized in computational pricing and large model inference, represent an effort to create standardized commercial exchange frameworks for data assets. Alongside commercial mechanisms, the government intends to improve the layout of academic disciplines related to data factors and roll out a specialized talent cultivation program for high-skilled data personnel.
The initiatives were detailed in conjunction with the release of the China Data Industry Development Report (2026) at the 2026 China International Big Data Industry Expo. The report indicates that the market scale of China's data industry has exceeded 6 trillion yuan. Products and services closely aligned with artificial intelligence now account for more than 70 percent of the sector's total output value.
The report also documented that China now has over 480,000 data-related enterprises, reflecting a rapidly growing commercial base. By combining centralized state-sector data releases with municipal-level data-labeling infrastructure, authorities are seeking to clear data bottlenecks that have constrained domestic generative AI deployment.
Sources
- 我国推出新举措加快数据产业发展 — State Council of China · 8/31/2026