Eight departments jointly issued the "Implementation Opinions on the Special Action of 'Artificial Intelligence + Manufacturing'."

Recently, eight departments including the Ministry of Industry and Information Technology issued the "Implementation Opinions on the Special Action 'Artificial Intelligence + Manufacturing'." The document sets forth the goal that by 2027, China will have achieved a secure and reliable supply of key core technologies in artificial intelligence, with its industrial scale and empowerment capabilities firmly ranking among the world's top. It also aims to promote the deep application of 3 to 5 general-purpose large models in manufacturing, develop specialized, fully-covered industry-specific large models, create 100 high-quality datasets for industrial applications, and popularize 500 typical application scenarios. Furthermore, it seeks to foster 2 to 3 globally influential ecosystem-leading enterprises, along with a group of specialized, refined, distinctive, and innovative small and medium-sized enterprises; cultivate a number of application service providers that are "well-versed in intelligence and deeply familiar with industries"; and identify 1,000 benchmark enterprises. By then, a world-leading open-source and open ecosystem will be established, with comprehensively enhanced security governance capabilities, thus contributing a Chinese solution to the development of artificial intelligence.

The Implementation Opinions point out:

(1) Strengthen the supply of AI computing power. Promote coordinated development of AI chip software and hardware, and support breakthroughs in key core technologies such as high-end training chips, edge-side inference chips, AI servers, high-speed interconnects, and intelligent computing cloud operating systems. Advance the orderly deployment of high-level intelligent computing facilities, accelerate the construction of computing power interconnection platforms and a nationwide integrated computing network monitoring and scheduling platform, launch pilot programs for intelligent computing cloud services, and promote the deployment of large-model all-in-one machines, edge computing servers, and industrial cloud computing resources, thereby enhancing the capacity to supply intelligent computing resources.

(2) Develop high-level industry-specific models. Support innovation in model training and inference methods, and develop high-performance algorithmic models that are tailored to the real-time requirements, reliability, and safety characteristics of manufacturing. Cultivate large-scale models for key industries, advance the development of a “cloud-edge-end” model architecture, and continuously enhance generalization capabilities. Create small-scale models geared toward specific industrial sub-scenarios, and encourage collaborative innovation between large and small models. Promote lightweight deployment of models and accelerate their practical application in industrial settings. Build a public-service platform for models, providing high-quality models and supporting tools.

(3) Launch the “Modular Resonance” initiative. Promote the establishment of a chief data officer system within enterprises, continue advancing the implementation of the national standard for data management capability maturity, and solidify the foundation of enterprise data governance. Compile a list of data resources that align with industry model requirements, issue guidelines for building high-quality datasets in manufacturing, and make effective use of platforms such as the Manufacturing Digital Transformation Promotion Center to facilitate the transformation of basic data into high-quality industry-specific datasets, thereby achieving “data-driven modeling.” Provide guidance to enterprises on strengthening their data engineering capabilities, promote deep integration between data development and model building, explore the establishment of an integrated mechanism encompassing “data collaboration, model training, application development, and security assurance,” and thus achieve “data-powered modeling.”

(4) Accelerate the empowerment of key industries through AI applications. Deeply carry out the “In-Depth Initiative” for AI-powered new industrialization, organizing high-level expert teams, enterprises, and research institutions to provide targeted support services that reach deep into industries, localities, and industrial parks. Build an AI application matchmaking platform to facilitate precise matching between supply and demand. Referring to the “Guidelines for AI-Powered Transformation of Key Manufacturing Industries” (see Appendix 1), develop industry-specific panoramic maps and transformation roadmaps for “AI + Manufacturing,” accelerate the empowerment of key manufacturing sectors—including raw materials, equipment manufacturing, consumer goods, electronic information, software, and IT services—and expedite the promotion and application of benchmark solutions and best practices.

(5) Accelerate the transformation and upgrading of the entire process. Systematically identify key application scenarios across all stages, deepen the tiered development of smart factories, and promote the deep integration of large-model technologies into core links of production and manufacturing. Transform the entire process—including R&D design (including industrial design), pilot testing and verification, production and manufacturing, marketing and service, and operations management—and enhance capabilities in areas such as assisted design, simulation model building, production scheduling and dispatching, and predictive maintenance of equipment.

— R&D and Design Phase: Focus on advancing intelligent assistive design, software code assistance tools, and drug discovery, thereby creating a new R&D and design model that is personalized, low-cost, and highly efficient. Strengthen the construction of industrial R&D datasets and promote open-source sharing; explore the establishment of an evaluation system for AI-predicted outcomes; enhance engineering and technological innovation capabilities; and clear the “bottleneck” hindering scientific breakthroughs in artificial intelligence.

— Pilot-scale verification stage. We will vigorously promote the intelligent transformation of pilot-scale operations, accelerate the application of technologies such as virtual simulation and multimodal fusion in the pilot-scale phase, and optimize process flows, enhance pilot-scale efficiency, and reduce experimental costs through comprehensive sensing, real-time analysis, scientific decision-making, and precise execution.

— Production and Manufacturing Stage: Deepen the application of artificial intelligence technologies in core industrial process control, process optimization, production scheduling, and other key areas, thereby promoting intelligent analysis, decision-making, and execution throughout the production process. Widely adopt industrial quality inspection technologies such as machine vision and unmanned intelligent inspections to strengthen real-time monitoring of production lines and predictive maintenance, enhance the accuracy of equipment fault detection, and achieve early warning of safety production risks and event alerts.

— Marketing service phase: Promote intelligent customer service, digital humans, and 3D product models. Focus on breakthroughs in personalized recommendations, customized after-sales services, and service-oriented extensions. Develop AI-based functions such as Q&A, training, and mediation to enhance the pre-sales, in-sales, and post-sales service experience and elevate service value.

— Operations management phase. Leverage the reasoning and prediction capabilities of large models to accelerate the intelligent upgrade of processes such as order processing, sales forecasting, and inventory alerts, thereby optimizing supply chain management. Utilize the large models’ analysis and generation capabilities to enhance enterprises’ management capabilities in areas including strategy, human resources, finance, and risk management.

(6) Enhance the application level of key enterprises. Conduct assessments of manufacturing enterprises’ readiness for AI applications and implement the “Guidelines for AI Applications in Manufacturing Enterprises” (see Appendix 2) to provide enterprises with implementation pathways and methodological guidance for intelligent transformation and upgrading. Encourage leading enterprises and central state-owned enterprises to take the lead in piloting and testing, providing large-scale application scenarios and proactively exploring new models and applications for AI-powered manufacturing. Deeply implement the special action to digitally empower small and medium-sized enterprises (SMEs), supporting them in carrying out digital and intelligent upgrades and accelerating the replication and promotion of AI applications among SMEs.

(7) Promote the application and popularization in key regions. Leverage the role of national pilot zones for innovative applications of artificial intelligence, build and open up a batch of “AI+Manufacturing” application scenarios, and create innovation hubs with distinctive industry characteristics. Relying on the advantages of national independent innovation demonstration zones and national high-tech zones—such as concentrated resources and a dense pool of talent—accelerate the large-scale implementation of new AI products, new services, and new business models. Support advanced manufacturing clusters and digital industry clusters to carry out AI-powered applications, and promote regional cluster-based transformation and upgrading.

(8) Promote intelligent upgrades in key areas. Strengthen the synergy between artificial intelligence and information and communication networks, foster the integration of AI with industrial internet platforms to enhance their capabilities, and develop datasets, large-scale models, and intelligent agents tailored for infrastructure such as the industrial internet. Advance the deep application of AI technologies throughout the entire lifecycle of infrastructure—including planning, construction, operation, and maintenance. Deepen the integration and application of AI technologies in the field of green manufacturing, developing and promoting intelligent, green, and collaborative solutions that address specific needs in areas such as energy and carbon emission management, and resource recycling. Create a batch of industry-specific application security solutions, accelerate the deployment of large-scale security models and intelligent agents, establish secure operational frameworks, and elevate safety standards in the industrial sector.

(9) Promote the iterative development of intelligent equipment. Accelerate the integration of artificial intelligence into industrial machine tools and industrial robots, develop next-generation AI-based CNC systems, and enhance capabilities in autonomous decision-making, analysis, and execution. Speed up the development of surgical robots and intelligent diagnostic systems, and accelerate innovation and clinical adoption of smart medical devices. Integrate AI technologies into the R&D, manufacturing, and operation of major technological equipment such as large aircraft and ships, and foster the development of intelligent low-altitude equipment like drones. Conduct testing and safety assessments of intelligent connected vehicle products equipped with autonomous driving functions, and steadily advance pilot programs for product approval and road trials.

(10) Accelerate the upgrade of smart terminals. Support technological breakthroughs in areas such as on-device models and development toolchains, and foster the growth of AI-enabled terminals including smartphones, computers, tablets, and smart home devices. Focus on key application scenarios such as industrial inspection and telemedicine, and speed up the industrialization and commercialization of new types of terminals—including augmented reality/virtual reality (AR/VR) wearable devices and brain-computer interfaces. Promote innovation in embodied intelligence products, establish pilot testing bases and training grounds for humanoid robots, build benchmark production lines for humanoid robots, and take the lead in deploying these robots in typical manufacturing scenarios.

(11) Foster new business models for software and intelligent agents. Accelerate the upgrading of traditional software products and services, expand software functionalities, and enhance user experience. Promote the deep integration of artificial intelligence with industrial software in areas such as R&D design, production control, and operational management, thereby boosting design and production efficiency. Conduct research and development on industrial intelligent agent technologies and promote their application, develop programming and development intelligent agents, and facilitate cloud-based deployment of intelligent agents. Develop open and collaborative intelligent agent protocols and interfaces to improve the interconnectivity and interoperability of intelligent agents. Select and showcase a number of exemplary cases of industrial intelligent agent applications, release enterprise-level application practice guidelines, and systematically advance the large-scale implementation of intelligent agents. Establish a classification and grading management system for intelligent agents, explore mechanisms for intelligent agent registration and discovery, identity authentication, and access management, and guide the healthy development of these new business models.

(12) Phased Cultivation of Enterprises. Support enterprises in increasing their investment in innovation, actively undertaking major national tasks, and pooling resources to build ecosystem-leading enterprises with global influence. Develop AI enterprise incubators, implement entrepreneurship support programs for small and medium-sized enterprises, and foster—through a phased approach—more specialized, refined, distinctive, and innovative “Little Giant” enterprises in the AI sector, high-tech enterprises, single-champion manufacturing enterprises, unicorn enterprises, and gazelle enterprises. Encourage relevant localities to provide enterprises with support measures such as “computing power vouchers” and “model vouchers,” strengthen public services that empower SMEs, and reduce the costs of enterprise development and application.

(13) Develop innovative platforms. Establish a National Manufacturing Innovation Center in the field of artificial intelligence to enhance the capacity for supplying key common technologies. Deploy a number of key laboratories in the AI domain and strengthen exploration of cutting-edge technologies such as brain-inspired intelligence and world models. High-quality build national pilot bases for AI applications in key manufacturing industries, pool industrial innovation resources, and accelerate the development of industry-specific solutions that are replicable and scalable.

(14) Develop empowerment application service providers. Improve the service system for digital transformation in manufacturing, establish a number of AI-powered application accelerators, cultivate high-quality empowerment application service providers, create standardized empowerment solutions, and offer services such as industry model optimization, data governance, and security assurance. Encourage industrial enterprises, AI companies, and industrial internet enterprises to pool resources—including tools, technologies, and platforms—to build ecosystem-based service providers. Support telecom operators and state-owned central enterprises’ digital and intelligent technology companies in enhancing their service capabilities and taking on industry-specific empowerment application services. Provide guidance to relevant industry organizations to regularly publish directories of high-quality service providers.

(15) Strengthen standard guidance. Leverage the roles of the Artificial Intelligence Standards Committee under the Ministry of Industry and Information Technology, the National Data Standardization Committee, the AI Sub-committee of the National Information Technology Standards Committee, the AI Chip Working Group of the National Integrated Circuit Standards Committee, and the Special Task Force on New Technology Security Standards of the National Cybersecurity Standards Committee to enhance the development of standardization technical organizations. Strengthen cross-industry and cross-sector collaboration, and promote—by tier and category—the formulation of foundational standards covering security, governance, and ethics; general standards integrating software and hardware; application-enabling standards; and metrological technical specifications. Establish a comprehensive evaluation system for key AI technologies, products, and enabling applications; support the development of an evaluation benchmark system for large-scale AI models; create authoritative evaluation rankings; regularly release evaluation results; and drive continuous technological iteration and upgrading. Deeply carry out the “AI Standards Initiative” activities to strengthen the dissemination and application of standards. Encourage enterprises to participate in international standardization efforts.

(16) Promote open-source and open collaboration. Build high-level, open-source AI communities, launch and implement a batch of open-source projects, and foster an AI open ecosystem with global influence. Organize developer conferences, “Campus Open Source Tours,” and other events to disseminate the open-source ethos and enrich open-source culture. Support the development of open-source tools such as development kits and model applications, and promote the practical application of open-source AI achievements.

(17) Strengthen talent attraction and development. Conduct forecasts of talent demand in the artificial intelligence industry, release reports on talent demand projections, and support universities and research institutions in proactively planning and adjusting and optimizing relevant disciplines and specialties. Build and effectively utilize the National Academy of Artificial Intelligence, the National Platform for Innovation in Industry-Education Integration in AI, the National Academy of Outstanding Engineers, and the National Base for Practical Training of Outstanding Engineers. Develop specialized courses to cultivate versatile talents who not only understand AI but also master its applications in manufacturing. Enhance AI-related cognitive education and training programs to improve the AI literacy and skills of all employees. Strengthen the cultivation of high-skilled talent in the field of AI, leveraging national talent programs and projects to nurture leading scientific and technological talents and innovative teams. Establish unconventional new models for cultivating leading talents and actively attract top overseas talent.

(18) Enhance security assurance capabilities. Focus on breakthroughs in key technologies such as deep synthesis forgery detection, security protection for industrial model algorithms, protection of training data, adversarial sample detection, and security assessment of smart terminals. Strengthen data security management and bolster the capacity for AI security protection. Build resource libraries—including security risk databases and corpus collections—and develop large-scale industrial safety models. Through knowledge base optimization, correction of training corpus errors, and identification of generated synthetic content, enhance the transparency and interpretability of AI systems and reduce the risk of hallucinations. Implement guidelines for the ethical management of AI technologies, strengthen industry self-regulation, and improve enterprises’ ability to prevent AI-related ethical risks.

(19) Establish a security governance mechanism. Study and develop security policy standards for the industrial and information technology sectors covering AI classification and grading, evaluation and testing, emergency response, and other related areas. Support local authorities in exploring flexible governance mechanisms. Build technical capabilities for monitoring and early warning of AI security risks, and strengthen risk monitoring, analysis, and prevention efforts. Formulate guidelines for reporting and sharing AI security risk information within the industrial and information technology sectors, coordinate resources across all links of the industrial chain, and enhance information sharing, risk notification, and collaborative response.

(20) Promote the high-level overseas expansion of advantageous industries. Encourage enterprises to tailor artificial intelligence products and empowering application solutions to the specific characteristics of different countries and regions. Launch an “overseas version” of the in-depth initiative to leverage AI for new industrialization, support industry organizations and specialized institutions in providing complementary services for enterprises’ overseas expansion, and guide enterprises to efficiently carry out various technical validations and compliance certifications, thereby better supporting the orderly overseas development of these industries.

(21) Build international cooperation platforms. Actively participate in discussions on AI-related issues under multilateral and bilateral cooperation mechanisms such as BRICS, SCO, ASEAN, G20, and APEC. Support the well-organized hosting of globally influential high-end competitions, exhibitions, and conferences, including the World Artificial Intelligence Conference and the Humanoid Robot Competition, and actively promote China’s benchmark cases in AI. High-quality establish the China-BRICS Center for AI Development and Cooperation, enhance the level of pragmatic cooperation, and foster synergistic development of global industries.