Companies in chips, machine tools and AI gain a top-level claim to lead applied state research projects, and state venture funds are told to back early-stage hard tech for longer.
China’s top Party and government offices have told the country’s research system to answer to industry. On October 21, the General Office of the CPC Central Committee and the General Office of the State Council jointly published a directive on fusing scientific innovation with industrial innovation, released in full by Xinhua as an authorized release and reprinted by China News Service. It runs to eight sections and 17 numbered measures. It names integrated circuits, industrial machine tools, high-end instruments, advanced materials and biomanufacturing as fields where core-technology work must reach "decisive breakthroughs" (决定性突破).
The document, titled the Opinions on Promoting the Deep Integration of Scientific and Technological Innovation and Industrial Innovation (关于推动科技创新和产业创新深度融合的意见), also orders the full implementation of the AI Plus (人工智能+) campaign so that AI is applied across all industries, and lists embodied intelligence, quantum technology and brain-computer interfaces among the frontier fields to be turned into future industries. Its organizing idea, as summarized in the Xinhua report carried by The Paper, is an innovation system led by enterprises (企业主导) in which companies, universities and institutes work together.
The directive is broad in scope and light on specifics. It sets no numeric targets, no budgets and no target years, and it names no specific regions for pilots. Its weight comes from who issued it: a joint Party and State Council directive sits above ministry plans, and the implementing rules, funding lines and local pilots that follow will draw their authority from it.
Companies are told to lead state projects and research is tied to commercial use
The most concrete change concerns who runs public research money. National science and technology projects with clear industrial application goals will in principle be led by enterprises, which will also steer how the results are commercialized. The directive also supports companies leading the construction of research platforms and bases, and it calls for coordination mechanisms that link research planning with industrial chains.
On finance, measure 11 tells government-funded venture capital funds and state-capital venture funds to invest earlier, in smaller companies, over longer periods and in hard technology, backed by longer appraisal cycles, differentiated supervision and better exit mechanisms. The same measure calls for using the national venture capital guidance fund (国家创业投资引导基金), building a "sci-tech board" (科技板) in the bond market to a high standard, and setting up a technology insurance policy system.
Governance gets its own lines in measure 14, on the market environment for innovation. It asks for laws and regulations in frontier fields such as AI and biotechnology to be improved, together with rules on technology ethics, and it calls for stronger supervision of new technology applications while exploring regulatory sandboxes, which let new products run under supervised trial conditions before full rules apply. Separately, measure 6 lists the frontier fields to be industrialized, running from quantum technology and biomanufacturing to hydrogen and fusion energy, brain-computer interfaces, embodied intelligence and sixth-generation mobile communications.
The directive adds an institutional layer above two earlier AI plans
AI Plus dates back more than a year. The State Council launched it in August 2025 with its own opinions, published on the website of the Cyberspace Administration of China (CAC). That document carried hard numbers: adoption of new-generation smart devices and AI agents above 70% by 2027 and above 90% by 2030. Earlier this month, the Party Central Committee and the State Council issued separate guidelines on "new quality productive forces" with 19 measures across five areas, which also called for fully implementing AI Plus, according to a Xinhua report published by ECNS. EastFrontier covered how those guidelines warned local officials against copycat AI investment.
The October 21 directive works at a different level. It leaves adoption goals to those earlier plans and rewires the machinery behind them: who leads projects, how research results move into products, how talent is trained and evaluated, and how venture and bond financing reaches technology companies. In this design AI Plus appears as one tool for upgrading traditional industries, listed alongside the industrial internet and the digital economy, rather than as the centerpiece. For chip and equipment companies, the language on integrated circuits and machine tools restates long-standing priorities, but it now comes paired with the instruction that enterprises lead the relevant national projects.
Ministries and local pilots will supply the numbers
The document closes by telling all regions and relevant departments to implement it under the Party Central Committee’s centralized leadership, adapting it to local conditions. Measure 17 asks international and regional science and technology innovation centers, and other qualified localities, to run reform pilots on deep integration, though it names none of them.
That leaves the substance to follow-on documents. The first places to look are ministry rules that define which national projects count as having clear industrial application goals (产业应用目标明确) and therefore must be led by companies, and any bond-market rules for the sci-tech board. A further marker would be appraisal rules for state venture funds that lengthen their review cycles, since that would change how quickly such funds need exits from early-stage chip, robotics and AI investments. With AI Plus already shifting toward scaling, provincial pilot lists under measure 17 will show which regions Beijing trusts to test the model first.