Implementation Opinions of the National Development and Reform Commission and the National Energy Administration on Promoting the High-Quality Development of “Artificial Intelligence+” Energy

Guo Neng Fa Ke Ji 〔2025〕 No. 73

To the development and reform commissions and energy administrations of all provinces, autonomous regions, province-level municipalities, cities with independent planning status, and the Xinjiang Production and Construction Corps, relevant central enterprises, and relevant industry associations:

In order to thoroughly implement the decisions and arrangements of the Party  Central Committee and the State Council on the development of artificial intelligence (AI), carry out the relevant work requirements set forth in the Opinions of the State Council on Deepening the Implementation of the “AI+” Initiative ([2025] Document No. 11), seize the major strategic opportunities in the development of AI, make our orientation toward applications more prominent, accelerate the deep integration of AI with the energy industry, and support high-quality development and high-level security in the energy sector, the following implementation opinions are hereby put forward.

I. Overall Requirements

Adhering to Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era as the guide, thoroughly implementing the spirit of the 20th Party Congress and the Second and Third Plenums of the 20th Central Committee, and fully implementing General Secretary Xi Jinping’s important instructions on promoting the deep integration of AI with the real economy and cultivating and expanding the AI industry, this work will rely on the expansion of application scenarios for the deep integration of AI with the energy sector as an important support, take the enhancement of the technological level of innovative applications of AI in the energy sector as its main focus, take the advancement of coordinated development between intelligent computing power (compute)  and electric power as a necessary support, and take the improvement of the innovation system for energy intelligentization development as a key safeguard. We will strive to enhance the safe, reliable, flexible, and efficient operational capacity of the energy system, ensure the secure and stable supply of energy and its green and low-carbon transformation, accelerate the cultivation of new quality productive forces, and provide strong support for the construction of a new-type energy system.

By 2027, a preliminary innovation system for integration of energy and AI will have been established, the foundation for coordinated development of compute and electric power will be further consolidated, significant breakthroughs will have been achieved in AI-empowered core technologies for energy, and applications will be more extensive and in-depth. Five or more specialized large models will be promoted for deep application in industries such as power grids, power generation, coal, and oil and gas. Ten or more key demonstration projects that are replicable, easy to promote, and competitive will be identified. Pathways for empowering 100 typical application scenarios will be explored. A number of research and development (R&D) innovation platforms for AI technology applications in the energy industry will be fostered. One hundred technical standards will be formulated and improved. A number of hybrid talents in energy and AI will be cultivated. Exploration will be carried out to establish a financial support system for research, development, and application of AI technology in the energy field. A development model for AI technology innovation in the energy sector that accords with China’s national conditions will be formed, and initial results in energy intelligentization will emerge.

By 2030, China’s dedicated AI technologies and applications in the energy sector will have generally reached a world-leading level. The mechanism for coordinated development of compute and electric power will be further improved, and an environmentally friendly, economical, safe, and efficient model of compute energy use will be established. Theoretical and technological innovation in the integration of energy and AI will have achieved notable results. AI technologies in the energy sector will empower cross-domain, cross-industry, and cross-business scenarios, achieving breakthroughs in areas such as intelligent power regulation, intelligent exploration of energy resources, and intelligent forecasting of new energy. Embodied AI and AI for science will achieve real-world applications in key scenarios. A number of globally leading “AI+” energy-related R&D innovation platforms and hybrid talent training bases will be established. A more complete policy system will be formed to continuously guide efficient, healthy, and orderly “AI+” energy innovation, laying a solid foundation for the high-quality development of energy.

II. Accelerating Empowerment of Energy Application Scenarios

1.AI+ power grid. Focusing on the requirements of power grid security, new energy integration, and operational efficiency under a new-type power system, carry out applications such as power supply–demand forecasting, intelligent grid diagnostic analysis, and intelligent generation of planning schemes in grid planning and design, and strengthen smart construction and management of power grid projects. Advance multi-scale intelligent simulation and analysis of power grids, explore the application of AI models in intelligent auxiliary decision-making and dispatch control of power grids, and enhance the safe, reliable, and low-carbon operation of all elements of source–grid–load–storage in the power system. Steadily improve the intelligentization level of R&D and manufacturing of key equipment such as transmission and substation facilities. Promote predictive maintenance of power equipment to prevent breakdowns, build intelligent agents for power equipment health management with autonomous sensing, decision-making, and execution capabilities, and enhance lean management of equipment. Advance integrated application of intelligent operation, distribution, and dispatching, build an intelligent support system for grid operation services, and improve the full-process intelligent service level for electricity customers. Promote the integration of AI technologies into the emergency power system and capacity building, and enhance the intelligentization level of power system disaster prevention, mitigation, and relief.

2.New business formats for AI+ new energy. Focusing on the needs of energy supply assurance and green, low-carbon transformation, advance the application of AI technologies in flexibly adjustable resources such as virtual power plants (including load aggregators), distributed energy storage, and vehicle-to-grid (V2G) for electric vehicles, in order to enhance load-side group control optimization and dynamic response capabilities. Strengthen the application of AI technologies in coordinated optimization and dispatching of new-type energy storage and power systems, as well as in total life cycle safety, and promote the intelligent optimization of renewable energy hydrogen production processes. Strengthen the empowerment of AI technologies in energy production processes for energy conservation and carbon emissions management, improve the total efficiency and carbon reduction level of multi-energy complementary integrated energy systems providing combined electricity, heat, cooling, and gas. Promote the application of AI in zero-carbon industrial parks, intelligent microgrids, and compute–power coordination, enhance the intelligent operational level of source–grid–load–storage integration, and facilitate local consumption of new energy.

3.AI+ new energy. In response to the volatility and intermittency of new energy output, accelerate the application of AI in areas such as high-precision power forecasting, electricity markets, smart operation of power plants, new energy planning, and post-project evaluation. Continuously promote the iteration and innovation of key materials and products for new energy. Advance the development of large models for power forecasting under complex scenarios and changeable weather toward smaller scales and higher precision, support wide-area coordinated optimization of new energy resources, and promote intelligent O&M of new energy power plants in remote areas. Build an integrated intelligent new energy production model of “meteorological forecasting + power forecasting + smart trading + intelligent O&M,” and make every effort to support a stable supply of new energy.

4.AI+ hydropower. Focusing on intelligentized construction of hydropower projects in high-altitude and cold regions and smart scheduling and operation of cascade hydropower stations, advance the application of AI technologies in hydropower project construction, and enhance the level of intelligentization of design, construction, and management of hydropower projects. Promote the integration of AI technologies with traditional hydrological models, meteorological models, and large-scale reservoir scheduling technologies, improve the accuracy of bidirectional coupled forecasting of meteorology and hydrology, and develop intelligent applications for optimization of scheduling decision-making. Promote the integration of technologies such as knowledge graphs, large models, and intelligent agents into a new-generation smart operation brain for hydropower, and form intelligentized solutions in key areas such as smart O&M and lean maintenance of hydropower stations, and intelligent dam situational awareness and smart management.

5.AI+ thermal power. Focusing on the development direction of cleanness and carbon reduction, safety and reliability, efficient regulation, and intelligent operation of thermal power, collaboratively carry out AI empowerment and technological innovation in operational scenarios such as fuel management and control, production operation optimization and intelligent control, and total life cycle management of equipment. Accelerate the digitalized design and construction and intelligentized upgrading of thermal power, promote the intelligentized development and application of thermal power operational control systems, and enhance the intelligent monitoring and health management capabilities for the total life cycle of key thermal power equipment, helping to further strengthen the support and assurance capacity of thermal power.

6.AI+ nuclear power. Focusing on the safe development of nuclear power, build intelligent support systems for nuclear power safety early warning, intelligent tracing and analysis of operating incidents at power plants, and emergency response. Conduct research on special O&M robot technology for the nuclear industry, continuously advance the technological upgrading and evolution of functions such as automatic startup and shutdown of nuclear power systems, explore technical pathways in which AI supports plasma prediction and control and controlled nuclear fusion, and promote the nuclear power industry’s steady transformation toward a new model that is data-driven, model-guided, and intelligently managed and controlled.

7.AI+ coal. Focusing on typical scenarios such as geological exploration, coal mine extraction (stripping), coal washing and processing, production scheduling, safety management and control, and equipment management, ensure stable acquisition of various operating data under complex geological conditions, multiple working conditions, and multitemporal and multispatial collaborative circumstances. Apply intelligent models in an integrated manner to achieve intelligent control and autonomous decision-making in the production process, support the normalized operation of reduced-personnel and unmanned work, steadily advance workforce reduction, safety enhancement, and efficiency improvement, and further consolidate coal’s role as a fundamental safeguard for energy security.

8.AI+ oil and gas. Focusing on directions such as cross-specialization collaborative research, field operation control, and production operation management and control, promote intelligent evaluation of geological targets for exploration, intelligent optimization of development plans, intelligent adjustment of drilling and fracturing operation parameters, intelligent operation of refining and chemical processing units, and real-time simulation of pipeline operation. Accelerate the R&D and application of intelligent production technologies and equipment such as intelligent drilling rigs, robots, UAVs, and intelligent sensing systems. Advance end-to-end intelligent linkage and automatic optimization at production sites, and promote the intelligentized upgrading and construction of the oil and gas production chain.

III.  Strengthening the Supply of Key Technologies

Focusing on technological bottlenecks in the energy sector such as data silos, fragmented compute, black-box algorithms, and high energy consumption of compute, promote R&D efforts on general purpose key technologies in data, compute, and algorithms applicable to the energy sector.

1.Consolidate the data foundation. In response to the need to build high-quality datasets and ensure data security in the energy sector, promote the application of technologies such as intelligent data labeling, intelligent enhancement, and data synthesis. Advance R&D on technologies such as energy data categorization and grading, privacy-preserving computing, intelligent dynamic data encryption, and cross-domain trusted traceability. Optimize data sharing mechanisms, accelerate the formation of high-quality datasets in the energy sector, and ensure the end-to-end security and reliability of energy data.

2.Strengthen compute support. In response to the need for integrated utilization of diverse heterogeneous compute under the combined model of leasing and construction in the energy sector, carry out R&D on key technologies such as unified scheduling of heterogeneous compute, intelligent orchestration of tasks, integrated storage-compute-network convergence, and compute pooling, so as to improve the level of intelligent computing services. Continue monitoring compute demand in the energy sector, coordinate planning of compute, electric power, and communication network resources, build a mechanism for coordinated development of compute and electric power through deep integration, and continuously increase the proportion of green electricity used by compute centers.

3.Enhance model foundation capabilities. In response to the energy sector’s demand for model security and explainability, advance the building of security capabilities for model algorithms and application systems, increase research on multi-agent collaboration, explainability, and lightweight model inference technologies, and continue to deepen applied research on key AI technologies in the energy sector such as machine vision, multimodality, and time-series forecasting. Promote deep integration of AI and software in the energy sector. To address the problem of high energy consumption in AI computing, accelerate breakthroughs in green and low-carbon AI technologies, research energy supply technologies such as flexible direct current electricity generation and small modular reactors, and encourage the application of efficient comprehensive energy utilization technologies such as liquid cooling in data centers, waste heat recovery, and intensive backup power.

IV.    Assurance Measures

1.Strengthen organization and implementation. Local main oversight departments for energy and relevant central enterprises must, in accordance with the requirements of these Opinions, establish and improve working mechanisms, coordinate and align relevant plans, and, in light of their actual conditions, accelerate the advancement of “AI+” energy development in their respective regions and work units. They must ensure the provision of all necessary factors of production, explore the establishment of a security governance system, form a working structure characterized by vertical coordination, layered implementation, and safe development, and accelerate the R&D, pilot demonstrations, and promotion and application for integrated AI application in the energy sector.

2.Promote collaborative innovation. Focusing on key general purpose technologies and complementary dedicated technologies for integrated innovation and application of AI in the energy sector, promote the establishment of a number of industry R&D and innovation platforms. Encourage enterprises to take the lead in joining with research institutions, universities, social service organizations, and other work units to establish cross-domain and interdisciplinary “AI+” energy innovation alliances aimed at technological innovation and integrated application. Deepen industry-academia-research-user cooperation, and build an open, collaborative, co-creative, and shared innovation ecosystem for energy intelligentization.

3.Strengthen development of standards and norms. On the basis of thoroughly summarizing application demonstration practices, accelerate the formulation of a set of technical standards and norms in areas such as energy data governance, integration of heterogeneous compute, and design of typical use cases. Promote the establishment of a standards system for AI in the energy sector, explore the creation of an evaluation indicator system for AI applications and an industry-level standards testing platform for AI applications, and improve the level of safe application of AI technologies in the energy sector. Encourage energy enterprises to take the lead in formulating international standards, and “go global” in technical standards to promote the application of AI technologies and products in overseas energy markets.

4.Carry out pilot demonstrations. Organize pilot demonstrations of AI applications in the energy sector, and select a group of scenarios and enterprise benchmark applications that are replicable and easy to promote. Encourage demonstrations of typical cross-domain and cross-industry scenarios, such as the integration of energy and transportation, and the integration of oil and gas with new energy. Prioritize including technical equipment related to AI applications in the energy sector within the support scope for the first units (sets) of major technical equipment in the energy sector. Support qualified regions and enterprises in carrying out pilot demonstrations of various AI applications in the energy sector based on local conditions, and conduct in-depth exploration and pioneering trials in areas such as technological innovation, commercial models, development formats, and institutions and mechanisms.

5.Increase support. Give full play to the stimulus role of central government funds, and, relying on national science and technology (S&T) major projects and key R&D programs in the energy and AI fields, steadily promote innovation in the application of AI technologies in the energy sector. Leverage the key hub role of multi-tier capital markets in supporting S&T innovation, and guide social capital to participate in the implementation of AI S&T projects and achievement conversion and application.

6.Improve the talent cultivation ecosystem. Encourage energy enterprises to jointly establish “AI+” energy talent cultivation bases with universities and research institutes. Design interdisciplinary curricula oriented to industry needs, with a focus on training hybrid talents equipped with knowledge of energy systems and the ability to apply AI algorithms. Increase the supply of such hybrid talents through industry–education collaboration.

National Development and Reform Commission and National Energy Administration

September 4, 2025