Digital News

In December 2023, the National Data Administration and 16 other government departments jointly issued the Three-Year Action Plan for “Data Elements ×” (2024–2026), which calls for application scenarios to drive the deeper integration and application of data elements across the economy and society. The “Data Elements ×” Competition, a key initiative under the action plan, has grown steadily since its launch in 2024. The number of competition tracks has expanded from 12 in the inaugural edition to 14 and then 17, while the number of participating teams has risen from approximately 19,000 in the first edition to more than 29,000 this year. In October 2026, in the final year of the three-year action plan, the national finals were held in Qingdao. Of the more than 29,000 teams that registered, 330 advanced to the final round of evaluation. For the first time, “data infrastructure” was included as a standalone competition track.

The competition is more than just a contest; it is also a barometer of progress in the market-oriented development of data elements. This year’s edition offers a window into how data elements are being harnessed to create value and how the data industry is changing. These developments not only demonstrate the tangible, verifiable results achieved under the three-year action plan, but also point to the direction and priorities of reform in the market-based allocation of data elements during the 15th Five-Year Plan period.

From Isolated Applications to Process Intelligence: Data Elements Become Deeply Embedded in Production

AI agents are emerging as a defining theme in industrial manufacturing. The most notable shift in this year’s industrial manufacturing track is that agents powered by large language models are moving beyond predictive maintenance for inpidual machines to enable coordinated decision-making across entire production lines. Among the projects that won first prize, Baosteel Zhanjiang Iron & Steel achieved real-time, closed-loop quality control throughout the production process; China National Petroleum Corporation (CNPC) completed intelligent optimization of shale oil reservoir stimulation; and China National Heavy Duty Truck Group (Sinotruk) developed a national-level demonstration project for intelligent commercial vehicle manufacturing. Together, these projects show how AI agents are moving from concept to real-world industrial applications, marking a shift from “data-driven control” to “intelligent manufacturing” as data elements become deeply embedded in production processes.

Data governance shifts from back-end support to front-line access control. For AI agents to operate reliably on production lines, data must be auditable, traceable, and accessible in real time. The maturity of data governance directly affects whether these agents can be deployed effectively in production environments. A common feature of this year’s winning projects is that auditable data provision, the systematic capture of process knowledge, and traceable decision-making have become baseline requirements. Data governance is no longer merely a back-office IT function. It has become a core operational capability that can determine whether an application succeeds or fails.

High-quality data supply is essential for scaling up applications. Across sectors ranging from energy equipment and display panel manufacturing to automotive production and steelmaking, the wider the range of applications, the greater the need for standardized data and rigorous quality control. “Data governance as a service” and other specialized service models are poised to become major growth areas. From data inventory and quality management to compliance audits, each stage of the process could create new markets for specialized services.

From a Technical Option to an Institutional Imperative: Trusted Data Circulation Becomes a Common Foundation Across Applications

Across financial services, healthcare, urban governance, and emergency management, trusted data circulation is emerging as a common theme. In financial services, trusted data circulation mechanisms are being put into practice in areas such as cross-departmental data sharing, risk management in inclusive finance, and new models of inclusive lending. In medical insurance, a trusted data space project led by the Big Data Center of the National Healthcare Security Administration is facilitating data sharing across institutions to support coordination among medical services, medical insurance, and pharmaceuticals. In urban governance, low-altitude sensing data and CIM-based spatiotemporal data are beginning to be integrated into the underlying digital infrastructure for urban management. In emergency management, coordinated systems that integrate multiple data sources have been successfully implemented in applications such as urban safety risk management.

What these projects have in common is not the volume of data involved, but how to ensure that data can flow freely, be effectively governed, and remain trustworthy across multiple stakeholders and geographical boundaries. Technical approaches such as trusted data spaces, privacy-enhancing computation, and data rights registration are being translated into concrete engineering practices that can be tested and validated in real-world settings.

From an Underlying Foundation to a Standalone Track: Data Infrastructure Enters a Phase of Systematic Development

For the first time, “data infrastructure” was given its own track in this year’s national finals, reflecting a growing industry consensus: without a high-quality foundation, higher-level applications cannot thrive. The projects included a computing power scheduling platform for a supercomputing network, data circulation and utilization infrastructure in the Yangtze River Delta, an integrated cross-regional network for multimodal data, and local nodes of the national data circulation and utilization infrastructure. Though perse in form, these projects share a common focus: building the infrastructure that enables data to flow, rather than developing applications that run on top of it. This points to a broader shift toward building out the infrastructure needed to facilitate the circulation of data elements.

The evaluation criteria from the infrastructure track are being extended to application-layer assessments. Capabilities such as trusted data circulation, data rights registration, and data valuation are becoming as important as application outcomes. Moreover, successful application scenarios rarely emerge from the efforts of a single organization. They are typically the result of collaboration between infrastructure operators and application service providers, pointing to an emerging ecosystem built on collaboration and mutual benefit across the data industry. For businesses, a practical priority is to identify their role in this ecosystem—as data providers, application scenario developers, or specialized service providers.

As the three-year action plan draws to a close, the market-based allocation of data elements is entering a new phase of accelerated development. The focus is shifting from data aggregation to data governance, from isolated applications to process-level intelligence, and from technological exploration to institutionalized mechanisms for trusted data circulation. Looking ahead to the 15th Five-Year Plan period, efforts to advance the market-based allocation of data elements are expected to deepen further in three key areas. First, China should accelerate the development of foundational data systems, moving arrangements for data property rights registration, data circulation and trading, and revenue distribution beyond pilot programs toward nationwide mutual recognition. Second, with infrastructure as the foundation, trusted data circulation as the safeguard, and high-quality data supply as a prerequisite, China should promote the large-scale realization of value from data elements across a wider range of industrial applications. Third, the practical experience gained from the 330 projects reaching the final evaluation stage across 17 competition tracks provides a concrete, evidence-based reference for the next phase of policy design and industrial development. The market-based allocation of data elements is moving from being merely usable to being genuinely useful, and from pilot programs to implementation at scale.

(By Bai Yang, Researcher at the People’s Data Research Institute)

Source: people.cn