In January 2026 the National Data Administration issued the Opinions on Strengthening Data Technology Innovation, setting out a systematic strategic path and implementation measures for data technology innovation. A landmark move in the document is that it defines the concept of “data technology” for the first time.
What “data technology” is
In the document’s terms, data technology is a systematic system of data science, technology and engineering built with the goal of giving play to data’s multiplier effect and releasing the value of data elements. The key to this definition is “systematic” and “engineered”: it elevates data from a “resource” to a “factor of production” that needs the support of a dedicated technology system.
Before this, data-related policies focused mainly on “circulation” and “rights confirmation”, such as data property rights registration and the cultivation of data-circulation service institutions. The Opinions place the emphasis on “technology”, meaning that building the data-element market has moved from a stage of institutional supply to a stage of capability supply.
Direct connection to AI development
The relationship between data technology and AI runs both ways. On one hand, the performance of AI models depends heavily on high-quality datasets; on the other, data technology itself is being reshaped by AI.
This is consistent with the National Data Administration’s existing programmes. The earlier action to build high-quality industry datasets, and the special programme on “data empowering artificial intelligence”, point to the same goal: enabling data to support model training in a usable, trustworthy and circulable form. See our notes on the action plan for high-quality industry datasets and the National Data Administration’s dataset guidance.
Implications for AI companies
First, data capability becomes assessable. With the concept of data technology established, “data governance level” gains clearer evaluation dimensions. In financing, bidding or cooperation, companies may be asked to explain the technical approach to data collection, processing and circulation.
Second, the path to data assetisation becomes clearer. Data property rights registration, the cultivation of data-circulation service institutions and data technology innovation form a complete chain; see the guidance on data property rights registration and our note on the opinion on cultivating data-circulation service institutions.
Third, compliance requirements at the circulation stage rise. Once the functional positioning of entities such as data traders and data-circulation platform enterprises is made explicit, their legality of data sources, authorisation chains and personal information protection duties will attract more attention. See our note on the network data security risk assessment measures and the PI outbound certification measures.
Fourth, cross-border data technology cooperation needs separate assessment. Technology cooperation involving outbound data should be judged according to the outbound route; see our cross-border data service.
Directions companies can start on
First, turn “dataset building” from a project task into an ongoing capability, clarifying data sources, labelling standards and quality assessment criteria. Second, map your own role in the data chain — data provider, processor or circulation service provider — and match duties accordingly. Third, follow data technology standards and pilots, accumulating technical documentation that can be relied upon in advance.
The introduction of the data technology concept essentially shifts the focus of competition in the data-element market from “who holds the data” to “who can turn data into usable assets”. For AI companies, this is both a capability requirement and a compliance requirement.
This article is based on the National Data Administration’s Opinions on Strengthening Data Technology Innovation and public interpretation. It is for general reference only and does not constitute legal advice.