Beijing Accelerates State-Led AI Mobilization Under Xi’s ‘New National System’

Executive Summary:

• Beijing is moving to systematically embed artificial intelligence into its national innovation system, according to a high-level leadership meeting of the Ministry of Industry and Information Technology (MIIT) in early June. This marks the beginning of a new phase in which AI development is treated as a system-wide strategic imperative.
• The Party now frames AI as a strategic national project, meaning quasi-private actors must align with state priorities to secure support. This is enhanced by new legislation, such as the Private Economy Promotion Law, and policy documents, such as the Intellectual Property Nation-Building Promotion Plan, which impose quasi-public obligations on firms and institutionalize state integration.
• MIIT outlined priorities including infrastructure upgrades, advancement of the AI technology stack, accelerated deployment of large models, establishment of technical standards and governance frameworks, and construction of integrated systems to drive AI development and innovation. These efforts blur boundaries across sectors and institutions, complicating the logic of targeted export controls as nearly any component may serve broader state-directed goals.

On June 3, the Ministry of Industry and Information Technology (MIIT) convened a leadership meeting to translate Xi Jinping’s April 25 Politburo directives on artificial intelligence (AI) into concrete implementation. MIIT Party secretary and minister Li Lecheng (李乐成) explicitly framed the session as a response to Xi’s “important instructions” (重要指示) and laid out a multi-pronged roadmap for embedding AI into the industrial system. This signaled the launch of a new phase of AI-industrial convergence (MIIT, June 4).

Xi’s instructions stated that AI development must fully exploit the “new national system” (新型举国体制), leveraging its advantages to uphold self-reliance and prioritize innovation (China Brief, May 21). By linking the new national system to AI, Xi signaled that this too is entering a high-priority, mobilized campaign—one that demands coordinated effort across ministries, industries, and quasi-private actors under Party-led strategic direction.

MIIT’s Implementation Blueprint: Mobilization Over Market

While the Politburo had elevated AI as a strategic imperative tied to national security and global competitiveness, MIIT’s focus was squarely on execution. This focus centered on four strategic pivots that collectively align with Xi’s repeated calls for full-stack domestic autonomy, AI-enabled industry, and institutionalized governance capacity, as well as a developed system to support these goals:

• Infrastructure and stack development: Strengthening computing power supply, coordinating general and industry-specific large models, advancing software-hardware integration, and building high-quality industrial datasets.
• Application-driven upgrading: Accelerated deployment of large models in manufacturing, refinement of real-world application scenarios, and cultivation of AI service providers to drive sectoral transformation.
• Standardization and governance: Hierarchical and systematic standards across AI domains, balancing market dynamics with strong state guidance.
• Ecosystem and incentive design: Support for specialized AI enterprises, such as expanded open-source platforms, improved fiscal and tax incentives, and guiding capital toward AI development.

Building Security into the Stack

The meeting also wove security governance into the industrial agenda. MIIT proposed embedding safeguards at every stage of AI development—from deep synthesis detection to ethics frameworks and technology risk oversight—treating risk as a core design principle. This approach directly reflected Xi’s call for early warning and response systems and underscored a broader strategy in which national security logic is inseparable from industrial modernization (China Brief, May 23).

In past domains such as COVID vaccine development, semiconductors, and lunar exploration, official sources (e.g. Xinhua, the State Council Information Office) attributed the PRC’s rapid progress to this centrally orchestrated system. These efforts did not focus solely on technological breakthroughs but also on downstream control—establishing standards, managing risk, and building institutional mechanisms to translate breakthroughs into long-term capacity. Now, with AI elevated to the same strategic tier, the same model is being extended: state-guided standards, platform-building, embedded risk controls, and targeted fiscal support are being mapped directly onto the sector.

Strategic Firms, State Obligations

The significance of harnessing AI to the “new national system” is twofold. First, it signals that AI is no longer merely a growth engine, but a multidimensional national project—one in which firms must align with Party-defined missions to access state support. Second, it implies that quasi-private actors like DeepSeek will increasingly operate under quasi-public obligations: sharing data, adhering to public procurement standards, and coordinating innovation pipelines with state priorities (China Brief, February 11). This governance approach is not simply about enabling industry. It is about synchronizing it with the Party’s long-term technological ambitions.

This synchronization is already being institutionalized through a suite of new laws and policy tools that extend state leverage over nominally private firms. During the April rollout of the Private Economy Promotion Law, state media singled out AI firm DeepSeek as an example of innovative “key players” eligible for strategic financing and policy support, highlighting the firm’s rise as emblematic of how private tech champions are being aligned with national priorities (Xinhua, April 30; NPC Observer, May 15). Access comes with conditions, however.

Under the 2025 Intellectual Property Nation‑Building Promotion Plan, firms engaged in major state projects must contribute proprietary IP to national pools and submit to centralized arbitration. This mechanism recalls the IP collectivization and procurement strategies of earlier campaigns like Made in China 2025 (State Council, May 19, 2015; CNIPA, May 9). For quasi-private firms, the new model requires operating within a governance framework defined by Party oversight, shared IP regimes, and policy alignment, in exchange for privileged access to capital, contracts, and state data. This is unambiguous state integration. Under the “new national system,” the PRC’s most promising AI applications are now industrial instruments within a broader architecture of centralized planning and Party-led coordination.

Conclusion

The PRC’s deep integration of AI into its national innovation system under the “new national system” makes it increasingly difficult for policymakers—especially in Washington—to draw clean lines between civilian and military, public and private, or commercial and strategic domains. Any one part of the system may directly or indirectly support another, meaning that AI compute infrastructure, open-source models, or enterprise applications can all serve state objectives. This systemic entanglement complicates the enforcement logic behind U.S. export controls, which rely on identifying specific end-users or downstream applications. As Beijing accelerates AI-industrial convergence through Party-led coordination, the very notion of a clearly defined “dual-use” boundary erodes—by design.

Just as important, the Party-state’s AI campaign is being built on increasingly routinized institutional architecture. The combination of governance protocols, incentive structures, and legal instruments—such as mandated IP pooling and capital allocation—suggests that the Party is building a replicable model for mission-driven innovation, one that subordinates market dynamics to long-term strategic goals. This shift toward managed ecosystems will not only reinforce domestic technological self-reliance but also give Beijing greater flexibility in insulating critical pathways from foreign disruption. In the years ahead, understanding the PRC’s AI trajectory will require less focus on individual firms and more on how systemic governance is being used to fuse innovation with strategic Party command.