China Academy of Information and Communications Technology (CAICT), through its Cloud Computing & Big Data Research Institute, has published the “Data Element Development Report (2025),” a research report that takes stock of how China’s data-factor agenda is evolving in practice. The report’s central framing is straightforward: data value is not unlocked by a single policy or platform, but by connecting an end-to-end chain—getting data “out” (supply), keeping it “moving” (circulation), making it “useful” (application), and keeping it “safe” (security).
A distinctive feature of the 2025 edition is its emphasis on the feedback loop between data and AI. The report argues that large-model development is intensifying demand for high-quality, multi-modal, and domain-specific datasets, while AI is simultaneously changing the economics of data governance by lowering the cost of processing and improving data quality through more automated, intelligent methods. It describes this as a reinforcing cycle—data enabling smarter systems, and smarter systems improving data.
What the report covers (and why it is useful):
1) “Supply”: Expanding high-quality data availability
The report reviews policy moves intended to increase the availability of usable data, including public-data development and the scaling of high-quality datasets. It highlights the growing push toward systematic dataset construction and the role of departmental and local coordination in turning policy goals into a pipeline of accessible, well-governed data resources.
2) “Circulation”: Building a more integrated data market
On circulation, the report spotlights the shift toward a “nationally integrated” approach to data markets—i.e., reducing fragmentation across regions and platforms and improving rule consistency. It also points to early momentum in data circulation infrastructure and references model contract templates as practical tools to reduce transaction and compliance friction for market participants.
3) “Application”: Moving from pilots to deeper, multi-layered use
The report treats application as the “last mile” of value creation. It notes continued emphasis on expanding data-enabled scenarios and on improving reuse across sectors. It also discusses the role of data-and-AI fusion in changing application paradigms, alongside ongoing exploration around data assetization and how economic value is recognized and managed in business settings.
4) “Security”: From compliance checklists to proactive, lifecycle governance
Security is framed not as a brake on circulation, but as the condition that makes scalable circulation viable. The report describes a shift away from basic compliance-only thinking toward more proactive governance that is integrated with business operations and covers the full data lifecycle, with increasing attention to more intelligent and precise security techniques.
A practical takeaway for readers outside China is that the report reads less like a single-topic “data policy” document and more like a systems view of how a data-factor strategy is operationalized: policy and standards development, infrastructure and market mechanisms, scenario-driven application, and security governance designed to scale with complexity—especially in an environment where AI development increases both the value of data and the risks of misuse.
