The Government Work Report submitted for deliberation on 5 March 2026 proposed, for the first time, “building a new form of intelligent economy”. It is the third consecutive year the report has deployed the “AI Plus” initiative, but the new formulation moves the focus from technology application to the economy itself.

The relevant passages include: deepening and expanding “AI Plus”, accelerating the adoption of next-generation intelligent terminals and agents, driving commercial and large-scale AI applications in key industries, cultivating AI-native business models; supporting AI open-source communities; implementing ultra-large-scale intelligent computing clusters and computing-power coordination as new infrastructure, and strengthening nationwide integrated computing scheduling; alongside improving the data-element system and AI governance.

Signal one: from technology breakthroughs to industrial integration

The phrase “new form of intelligent economy” implies that AI’s role is shifting from an imported tool to infrastructure of the economy. Correspondingly, policy resources shift from single-point technology to industry deployment and scale.

The practical implication: being able to explain “in which industry scenario you deliver at scale” matters more than explaining “how strong your model is”. This orientation is already visible in sector-specific action plans — see our note on MIIT’s special action to cultivate AI application service providers.

Signal two: data and compute are framed as parallel foundations

The report places “deepening the development and use of data resources” in the same passage as “nationwide integrated computing power”. Policy has effectively treated data and compute as the two foundational supply lines of the intelligent economy.

Institutional supply on the data side is accelerating. The National Data Administration’s push on high-quality industry datasets is consistent with this direction — see our analysis of the action plan for high-quality industrial datasets; for the annual data-element agenda, see our note on the 2026 digital economy work points.

For AI companies, the barrier to participating in dataset building is not technology but lawful data sources and personal-information boundaries — a point made more explicit in the Model-Data Resonance action.

Signal three: development and governance are deployed together

The report addresses both scaling commercial applications and “improving AI governance” in the same passage. This is not rhetoric but sequencing: governance will advance in step with application expansion rather than waiting for problems to accumulate.

Companies should expect that as applications scale, enforcement of existing duties — algorithm filing, synthetic-content labelling, security assessment — will intensify. For the governance framework, see our note on the State Council’s AI development and deployment arrangements.

Three things companies can do now

First, put “industry scenario plus a scaling path” into external materials — it serves both financing narrative and policy fit.

Second, build a data-asset ledger. The data-element system will bring new tools for registration, circulation and authorisation, and companies with ledgers will connect more easily — see our note on the draft guide on data property-rights registration.

Third, bring governance duties into the product roadmap. Compliance is no longer a pre-launch patch but a design constraint running alongside feature iteration.

“Intelligent economy” is a long-term narrative. For AI companies, it means a narrative shift from “describing capability” to “describing delivery and compliance”.

This article is based on the public content of the Government Work Report submitted for deliberation on 5 March 2026. It is for general reference only and does not constitute legal advice.