The Central Economic Work Conference was held in Beijing on 10 and 11 December 2025. In setting out next year’s priority tasks, it called for “deepening and expanding ‘AI plus’ and improving AI governance”.
Two actions in one sentence
The phrase contains two parallel actions worth reading separately.
“Deepen and expand ‘AI plus’” is the direction of development. It stresses not whether to use AI but how to embed it in more industries. This echoes “fully implement the AI plus action” in the 15th Five-Year Plan proposals, showing “AI plus” has risen from a sectoral policy to a cross-cycle industrial direction.
“Improve AI governance” is the accompanying requirement. Development and governance appear in the same sentence, meaning the orientation is not “develop first, govern later” but simultaneous advancement. For companies, this means policy support comes with corresponding governance duties.
Connecting to existing policy
“AI plus” is not a new concept but a policy set of sectoral actions. Ministries including the MIIT have issued AI plus implementation opinions for manufacturing, consumption, software, transportation and employment services, forming a policy matrix from technology to scenarios.
The Central Economic Work Conference statement is the top-level confirmation of these sectoral actions — it gathers dispersed industry policies into a single development direction. See our notes on AI plus manufacturing, AI plus consumption, AI plus transportation and AI plus employment services.
Three signals for AI companies
First, industry scenarios remain the main arena for policy support. Policy resources will keep tilting toward “AI combined with specific industries”. Packaging capability into landable industry solutions connects more easily with support than emphasising general model capability.
Second, “improving governance” means compliance is a threshold, not a cost. When governance is written into the top-level economic work deployment, compliance capability becomes an implicit threshold for participating in industry projects. In procurement by government, state-owned and large enterprises, the completeness of algorithm filing, data compliance and security assessment often bears directly on winning. See our algorithm filing service.
Third, policy windows and compliance rhythm must be planned together. Sectoral actions usually set stage targets and timetables; companies seeking entry into supported scenarios need to complete product and compliance preparation within those timetables.
An observation
Placing “development” and “governance” side by side in priority tasks is a consistent style in recent policy statements. The message is clear: the AI industry has substantial room to grow, but the way it grows needs to be governable and traceable. For startups, treating governance capability as part of the product rather than a formality to be completed afterwards fits this orientation better.
This article is based on public reporting of the December 2025 Central Economic Work Conference and existing policy documents. It is for general reference only and does not constitute legal advice.