National Data Administration: By the end of 2025, more than 100,000 high-quality datasets will have been built nationwide, roughly 310 times the total digital resources of the China National Library.

On March 24, the State Council Information Office held a press conference to introduce details about the 9th Digital China Construction Summit.

Liu Liehong, Director of the National Data Bureau, stated on-site that by March of this year, China’s daily Token (word unit) invocation volume has exceeded 140 trillion. This is more than 1,000 times the 100 billion at the beginning of 2024, and in just three months, it has grown by over 40% compared to the 100 trillion expected by the end of 2025. The significant increase in daily Token invocation volume clearly indicates that China’s artificial intelligence development has entered a rapid growth phase.

Image source: Photo taken on-site by Jiayi Zhou, reporter for Daily Economic News

By March this year, China’s daily Token invocation volume has increased more than 1,000 times from 100 billion at the start of 2024.

Daily intelligent assistants, industrial intelligent analysis, and other applications rely on massive high-quality data support. What work has the National Data Bureau done to promote high-quality datasets to empower AI development, and what are the next steps?

Liu Liehong said that the National Data Bureau attaches great importance to data elements empowering AI innovation and development. To address the issues of “small and scattered” high-quality dataset construction, it has organized 26 departments to select 72 leading units for high-quality dataset construction, 140 pilot units, and 104 typical cases. They have built an ecosystem for high-quality dataset development driven by leading entities, multi-party participation, joint efforts, shared resources, and win-win cooperation, continuously promoting high-quality dataset creation.

To promote the development of the data annotation industry, the bureau has deployed seven cities—Chengdu, Shenyang, Hefei, Changsha, Haikou, Baoding, and Datong—to undertake pilot projects for data annotation. It issued the “Implementation Opinions on Promoting High-Quality Development of the Data Annotation Industry,” selected 47 excellent data annotation cases, and organized seven supply-demand matchmaking events for data annotation. Next, the bureau will focus on regions with strong technological innovation, solid development foundations, and distinctive industrial features, concentrating on “knowledge-intensive” and “technology-driven” directions, and will gradually establish a series of technologically advanced, distinctive, and efficient data annotation industry innovation pilot zones.

Liu further pointed out that the bureau is also continuously cultivating a market consensus that “pay for high-quality data,” promoting high-quality industry datasets to be listed, shelved, and traded on data exchanges. It supports data circulation service platforms and data vendors in providing trading services, encourages various data circulation service organizations to explore diverse models for high-quality data flow and utilization, and promotes orderly supply and demand matching of high-quality datasets, facilitating the free flow of high-quality data.

China’s high-quality dataset construction has achieved phased results. By the end of 2025, over 100,000 high-quality datasets will have been established nationwide, with a total volume exceeding 890 petabytes (PB), roughly 310 times the total digital resources of the National Library of China. As of March this year, China’s daily Token invocation volume has surpassed 140 trillion, more than 1,000 times the volume at the beginning of 2024, and in just three months, it has grown by over 40% compared to the 100 trillion expected by the end of 2025.

“The massive increase in daily Token invocation volume fully demonstrates that China’s AI development has entered a stage of rapid growth. Application scenarios are continuously deepening—from conversational AI to decision-making and execution agents. The competitiveness of China’s AI industry has also significantly strengthened. The current hot topic of Token going overseas is a sign of this industry competitiveness. From a data perspective, it also indicates that data supply is increasing substantially, and the value of data elements is being continuously released. The empowerment of data elements for AI innovation and development has entered a virtuous interactive stage,” Liu explained.

Liu emphasized that the next step will be to continue promoting data-driven AI innovation, working with all parties to implement a new round of high-quality dataset construction plans. These include six major initiatives: strengthening infrastructure and capacity expansion, tackling annotation challenges, improving quality and efficiency, enabling applications, managing services, and unlocking value. Guided by scenario needs, efforts will be accelerated to advance pilot projects, aiming to create AI-Ready high-quality datasets that are technically feasible, practical, and reliable, thereby increasing both the quantity and quality of high-quality data supply.

Promoting Policies for Data Elements to Enable New Industrialization

The Daily Economic News also noted that recently, the Ministry of Industry and Information Technology issued a notice to launch the Industrial Data Foundation Action, initiating pilot projects for high-quality industry datasets to empower AI. How will further progress be made?

Wang Yanqing, Director of the Department of Information Technology Development at the Ministry, said that moving forward, the ministry will continue to focus on three areas. First, strengthening support and guarantees. It will collaborate with local industry and information technology departments and data authorities to ensure resource support and guidance for pilot units, promptly address issues encountered, gather experience, and accelerate the development of scalable, promotable results.

Second, reinforcing policy guidance. It will promote the issuance of policies for data elements to enable new industrialization, release reference guidelines for industrial scenario data application, and strengthen development guidance and model promotion.

Third, cultivating a healthy ecosystem. It will accelerate the development of industrial data standards, grow data consulting, governance, and annotation enterprises, support the organization of technical seminars and supply-demand matchmaking events, and enhance the AI open-source community to create a high-quality open data resource hub. Notably, at the upcoming summit, the Ministry will host a special session on data elements empowering new industrialization and invite representatives from pilot units to share experiences. Additionally, a 2026 competition on data elements enabling new industrialization will be launched.