A new round of scientific and technological revolution and industrial transformation is reshaping the global competitive landscape with unprecedented depth. As the core carrier for seizing the commanding heights of science and technology, cultivating new quality productive forces, and building a modernized industrial system, future industries have become a focal area in the strategic game among major powers. During the 24th collective study session of the Political Bureau of the CPC Central Committee, General Secretary Xi Jinping explicitly pointed out: "We must stand at the strategic height of promoting the construction of a strong country and the great rejuvenation of the Chinese nation, base ourselves on objective conditions, leverage comparative advantages, adhere to the principle of pursuing progress while ensuring stability, and cultivate industries in a phased manner, so as to continuously achieve new breakthroughs in the development of China's future industries." This important discourse provides fundamental guidance for the strategic layout, development path, and policy practice of China's future industries. Under the guidance of national top-level design, the Ministry of Industry and Information Technology and other departments have systematically laid out six major industrial systems: future manufacturing, future information, future materials, future energy, future space, and future health. The cultivation of China's future industries has gradually entered a critical stage of systematic promotion, large-scale breakthroughs, and ecological growth, moving beyond early conceptual exploration and localized pilot programs. In this process, how to grasp the law of technological transition, optimize industrial organization methods, improve the allocation mechanism of innovation resources, and enhance the adaptability of the governance system has become a crucial factor determining the development quality of future industries.
Currently, a new generation of AI technologies, represented by Large Language Models (LLMs), AI for Science, and autonomous agents, is accelerating its evolution. The development paradigm of AI is expanding from "perception and recognition, assisted cognition" to "autonomous decision-making, closed-loop execution, continuous learning, and self-evolution." Its functional positioning is also undergoing profound changes: it is no longer merely a technical tool to improve production efficiency, but increasingly an important infrastructure supporting knowledge discovery, technological innovation, industrial organization, and resource allocation; it is no longer just an intelligent system for assisted decision-making, but gradually growing into a key participant in the innovation process, restructuring the innovation chain, and collaboratively driving industrial transformation. Thus, AI is profoundly changing the underlying logic of cultivating and developing future industries.
Unlike the linear upgrading logic of traditional industries, future industries are characterized by foresight, strategic importance, and disruptiveness. Their cultivation and expansion cannot be achieved solely through single technological breakthroughs, individual policy support, or efforts in a single field. Instead, it is a complex systems engineering involving the coordinated evolution of the technological innovation system, resource allocation system, industrial organization system, and governance system. As the core driving force of the new round of scientific and technological revolution, AI is driving future industries from single-point innovation competition to innovation ecosystem competition, from factor input-driven to data intelligence-driven, and from industrial chain competition to innovation network competition. Based on this, this article, grounded in the actual development of China's future industries and closely following the frontier trends of AI empowering industrial transformation, analyzes the paradigm shift and realistic bottlenecks in cultivating future industries, and proposes an integrated cultivation path centered on AI, providing a reference for promoting the high-quality development of China's future industries.
01AI Drives Systematic Transformation of the Cultivation Paradigm for Future Industries
Future industries are the product of the deep integration of cutting-edge technologies, emerging factors, innovative organizations, and new governance models. Their cultivation is essentially a systematic leap from the traditional industrial development paradigm to the new quality productive forces-driven paradigm. The deep penetration of AI technology is promoting comprehensive and profound changes in future industries, from technological logic to factor logic, and from organizational logic to value logic.
Technological Logic: From Technology Innovation-Driven to Science Discovery-Driven
Traditional industrial innovation mostly follows the linear path of "application traction and technology follow-up," focusing on the industrialization of mature technologies. In contrast, the innovative development of future industries relies more on basic scientific breakthroughs, cross-disciplinary integration of cutting-edge technologies, and the traction of disruptive application scenarios. It presents the significant characteristics of deep coupling between scientific discovery and technological innovation, and two-way interaction between basic research and industrial applications. Therefore, the cultivation of future industries must highlight the leading role of basic research, the strategic layout of cutting-edge technologies, and the application verification of disruptive technologies. The breakthroughs in AI technology are driving a fundamental shift in the paradigm of scientific research, becoming a key bridge connecting basic research and industrial applications.
The new scientific research paradigm represented by AI for Science is profoundly changing the underlying logic of technological cultivation in future industries. AI is no longer just an auxiliary computing tool in scientific research activities, but is gradually becoming an important intelligent system participating in knowledge discovery, pattern recognition, experimental design, and technology verification. Through large-scale data analysis, complex model deduction, automated experimental design, and high-throughput simulation verification, AI promotes the transformation of scientific research innovation from "experience-driven and hypothesis-driven" to "data-driven, model-driven, and intelligence-driven," thereby significantly improving the efficiency of scientific discovery and the possibility of technological breakthroughs.
In the field of quantum technology, AI can be used to optimize qubit control, improve quantum circuit design, and enhance quantum error correction efficiency, thereby accelerating the evolution of quantum computing from laboratory prototypes to practical systems. In the field of biomanufacturing, AI can deeply participate in key links such as target discovery, protein structure prediction, molecular design, drug screening, and experimental verification, greatly shortening the drug R&D cycle. In the field of advanced materials, AI can achieve directional design and rapid verification of high-performance materials through material performance prediction, structure optimization, and virtual screening, significantly reducing the cost of traditional "trial-and-error" R&D.
It can be seen that the integrated innovation model of "AI + cutting-edge science" is opening up the entire chain from basic theoretical breakthroughs to key technological research, and from prototype verification to industrialization applications. Its core significance lies not only in improving R&D efficiency but also in reconstructing the knowledge production mode and technology generation mechanism, shifting the cultivation of future industries from relying on single-point technological breakthroughs to systematic innovation based on AI for Science platforms.
Factor Logic: From Traditional Factor-Driven to Data Intelligence-Driven
The production factor system of future industries has undergone a fundamental reconstruction. Traditional factors such as land, labor, and capital still play a basic role, but their marginal contribution has relatively declined. New factors such as data, algorithms, computing power, and high-level talents have become key variables determining the innovation capability, development speed, and competitive advantage of future industries. As a brand-new techno-economic system, AI has the huge potential to digitally reorganize and intelligently substitute traditional production factors. By connecting data, algorithms, and computing power, it builds a "data-model-computing power" iterative system to achieve optimal allocation of innovation resources and continuous improvement of innovation efficiency.
As the core production factor, data is the "core raw material" for future industrial innovation. Its scale, quality, circulation efficiency, and development and utilization level directly determine industrial innovation capability. AI can enhance the intelligent level of data collection, cleaning, labeling, governance, rights confirmation, circulation, and trading, promote the orderly flow of data resources across entities, industries, and regions, and break the constraints of "data silos" and "data barriers" on future industrial innovation.
As the core hub of resource allocation, algorithms determine the depth of data mining, the efficiency of model reasoning, and the accuracy of scenario adaptation. With the development of large models, reinforcement learning, multimodal learning, and agent technologies, algorithms are evolving from single-task optimization to general capability generation, complex task reasoning, and autonomous iterative optimization, providing more flexible, efficient, and transferable intelligent support for the development of future industries.
Computing power is the basic support for innovation activities. High-complexity tasks such as large model training, quantum simulation, biological computing, industrial simulation, and digital twins highly depend on large-scale, high-performance, and low-cost computing power supply. AI technology can further promote the intelligent scheduling, elastic allocation, and efficient utilization of computing power resources, driving the development of computing power infrastructure from decentralized construction to national integration, networking, and universal accessibility.
As the primary resource, talents are the core force for original innovation, technological research, and ecosystem construction. Unlike traditional industries that require single-skilled talents, future industries rely more on compound talents with scientific literacy, engineering capability, data thinking, and industrial understanding. AI can not only assist researchers in improving the efficiency of knowledge acquisition, experimental design, and technological research but also improve the quality of talent cultivation through intelligent education, virtual experiments, and digital tutors, accelerating the formation of a high-level talent system adapted to the development of future industries.
Organizational Logic: From Human-Led Innovation to Human-Machine Collaborative Innovation
Future industries are characterized by high cross-boundary integration. A single enterprise, discipline, or technological field is difficult to independently complete the entire chain of innovation from basic research, technological research, engineering verification, to market application. Therefore, the cultivation of future industries requires a more open, collaborative, and networked innovation organization form. The development of AI, especially autonomous agent technology, is driving the innovation organization model from traditional hierarchical, closed, and linear division of labor to open collaboration, distributed innovation, and human-machine collaboration.
Agents possess capabilities such as environmental perception, autonomous planning, task decomposition, tool invocation, feedback learning, and continuous optimization. They can undertake functions such as information analysis, solution generation, path deduction, risk identification, and execution feedback in complex innovation tasks. In the process of scientific research innovation, they can assist in completing literature review, research hypothesis generation, experimental design, data analysis, simulation verification, and achievement optimization, improving the systematicness and efficiency of scientific research. In the process of industrial innovation, they can participate in product design, process optimization, supply chain management, market forecasting, operational decision-making, and customer service, achieving intelligent support for the entire process of industrial innovation activities.
It can be seen that future industrial innovation will gradually form a new collaborative model of "scientist + agent," "engineer + agent," "entrepreneur + agent," and "manager + agent." Innovation subjects are no longer limited to human individuals or organizations but are expanded into a human-machine collaborative innovation community composed of human experts, intelligent systems, digital platforms, and industrial organizations. Under this model, humans mainly undertake high-level functions such as value judgment, strategic choice, complex problem definition, and institutional design, while AI plays an important role in knowledge processing, solution generation, process optimization, and real-time feedback. Meanwhile, the development of open-source AI platforms, model toolchains, standardized interfaces, and cross-system communication protocols further reduces the collaboration cost among innovation subjects, promotes the efficient flow of knowledge, data, tools, and capabilities among different subjects, supports real-time collaboration across regions, disciplines, and organizational teams, and forms a more flexible, efficient, and open innovation network.
Value Logic: From Industrial Chain Competition to Innovation Ecosystem Competition
Value creation in future industries is no longer limited to the production and sales of single products but is the comprehensive output of technological capabilities, solutions, service systems, data value, and ecological value. As the platformization and open-source trends of AI platforms continue to strengthen, innovation activities are breaking through organizational and geographical boundaries. The competition in future industries is no longer limited to industrial chain competition but has become an ecosystem competition among innovation platforms, open-source communities, data resources, scenario systems, and innovation networks.
Value creation in traditional industries is mainly concentrated in the production and manufacturing links. Enterprises usually participate in industrial division of labor as product suppliers, and the value realization method is mainly one-time product transactions. Future industries emphasize the entire chain of value creation including R&D, design, manufacturing, service, operation, feedback, and iteration. The role of enterprises has also changed from pure product providers to technology service providers, capability exporters, system integrators, and ecosystem co-builders. For example, in the field of embodied AI, the value of enterprises is not limited to selling hardware products such as humanoid robots, intelligent equipment, and sensors, but is also reflected in providing customized solutions, remote operation and maintenance services, continuous software upgrades, and data closed-loop optimization for scenarios such as industrial manufacturing, medical care and rehabilitation, commercial services, and special operations. In the field of biomanufacturing, enterprises are no longer just producers of traditional pharmaceutical or biological products but provide full-process technical services and platform capability output around target R&D, molecular design, process optimization, pilot scale-up, and large-scale production.
By deeply embedding into the entire process of R&D, production, service, and operation, AI makes the value creation process more perceptible, computable, traceable, and optimizable. On the one hand, AI can enhance enterprises' real-time identification capabilities of user needs, scenario changes, and technological bottlenecks, driving value creation from supply-oriented to demand-oriented and scenario-oriented. On the other hand, AI can also support continuous iteration based on data feedback, shifting products and services from static delivery to dynamic optimization. On this basis, the value realization method of future industries will also undergo profound changes. The traditional "one-time transaction" model will gradually evolve into new business models such as "long-term co-creation, continuous service, pay-for-performance, and value sharing."
In summary, AI is driving a systematic transformation of the cultivation paradigm for future industries from four dimensions: technological logic, factor logic, organizational logic, and value logic. Its profound impact is reflected in the overall reconstruction of the scientific discovery mode, resource allocation mechanism, innovation organizational structure, and industrial value creation model. The competition in future industries is essentially an innovation ecosystem competition with AI as the core support, and a comprehensive competition of new quality productive forces generation capability, organizational collaboration capability, and institutional adaptability.
02Cultivation of China's Future Industries Enters a Critical Period of AI Empowerment
Under the guidance of national strategy, continuous policy efforts, rapid technological iteration, and market demand pull, the cultivation of China's future industries has achieved phased results. As the core engine of the new round of scientific and technological revolution, AI's capabilities continue to strengthen across the entire chain of R&D, manufacturing, service, and governance, injecting strong momentum into the development of future industries. However, at the same time, bottlenecks such as insufficient original innovation, reliance on others for core technologies, unsmooth factor circulation, weak ecological collaboration, and lagging governance adaptation still exist, restricting the full release of AI empowerment efficiency.
Development Achievements: Multi-Dimensional Breakthroughs, Continuously Consolidating the Foundation for AI Empowerment
1. Systematic improvement of top-level design and comprehensive consolidation of strategic consensus. The CPC Central Committee and the State Council have placed future industries at the height of the overall national strategy. The "15th Five-Year Plan" explicitly proposes forward-looking layout of future industries, focusing on six key areas. The Ministry of Industry and Information Technology and six other departments jointly issued the "Implementation Opinions on Promoting the Innovative Development of Future Industries," systematically laying out the industrial development in six directions. The State Council issued the "Opinions on Deeply Implementing the 'AI+' Action," explicitly proposing quantitative targets for the relevant intelligent penetration rate in 2027 and 2030, providing clear guidance for the digital and intelligent development of future industries. Local governments have responded actively. As of now, about 20 provinces across the country have issued special plans for future industries. Provinces and municipalities such as Beijing, Shanghai, Guangdong, and Jiangsu have built a batch of future industry pilot zones, forming a good cultivation pattern of "national coordination, departmental collaboration, and local efforts."
2. Steady leap in innovation capability, forming multi-point breakthroughs in AI empowerment. China's comprehensive innovation strength remains at the forefront of the world. In 2025, the R&D investment intensity of the whole society reached 2.8%, exceeding the average level of OECD countries. The proportion of basic research funding broke through 7%. The number of papers in high-level international journals, international patent applications, and the total number of R&D personnel have ranked first in the world for many consecutive years. In key areas of future industries, China has achieved multi-point breakthroughs: the number of patents in the field of quantum communication accounts for more than half of the global total, and a 300 km-level fully connected quantum direct communication network has been built; in the field of humanoid robots, a full industrial chain manufacturing capability has been formed, breaking through key technologies such as high-precision control and flexible movement; in the field of brain-computer interfaces, invasive technology has entered the clinical trial stage; in the field of 6G, key technology R&D is accelerating, and multiple achievements have reached the international advanced level.
At the same time, AI technology has injected strong momentum into the improvement of innovation capabilities. Domestic large models, intelligent toolchains, and agent tool platforms are iterating rapidly. They have become important tools for promoting innovation breakthroughs in fields such as biomanufacturing, new material R&D, and industrial software, providing low-cost and high-performance innovation support for future industries, greatly reducing the R&D threshold and accelerating the transformation of achievements.
3. Accelerated formation of the industrial ecosystem, with the enterprise echelon continuously expanding. China has the world's most complete, largest-scale, and most resilient industrial system. The scale of the manufacturing industry has ranked first in the world for 16 consecutive years, providing the industrial soil of full-chain support, low-cost trial and error, and rapid iterative optimization for future industries. The echelon cultivation of enterprises has achieved remarkable results. By the end of 2025, there were 509,000 high-tech enterprises nationwide, and over 17,000 "Little Giant" enterprises with specialization, refinement, differentiation, and innovation. A batch of unicorn enterprises and manufacturing single-champion enterprises have emerged in future industry fields such as quantum, biology, and AI. The deep integration of the industrial chain, innovation chain, capital chain, and talent chain has initially formed an open and collaborative industrial ecosystem.
4. Continuous expansion of application scenarios, accelerating the transformation of market advantages. China has a population of over 1.4 billion, more than 400 million people in the middle-income group, and over 180 million market entities. The super-large market provides massive application scenarios and broad development space for future industries. The integration of future industries and the real economy continues to deepen: in the manufacturing field, industrial AI and embodied AI empower the construction of smart factories, driving the intelligentization of the entire production process; in the service field, medical AI, educational AI, and financial AI are implemented on a large scale, significantly improving service efficiency and inclusiveness; in the agricultural field, digital planting, precision breeding, and intelligent agricultural machinery are accelerated to be promoted, effectively driving agricultural modernization. In 2025, the scale of China's core AI industry reached 1.2 trillion yuan, providing solid market support for the cultivation of future industries.
Prominent Bottlenecks: Multiple Constraints, AI Empowerment Efficiency Needs to be Improved
1. Weak original innovation capability, core technologies restricted by others. China's future industries as a whole are still in the stage of following and partially running alongside others. The output of disruptive original innovation is insufficient. Although the intensity of basic research investment (7.1%) continues to increase, it is still significantly lower than that of the United States (15.4%) and Japan (12.3%). At the hardware level, high-end AI chips with processes below 7 nm highly depend on imports, and the performance of domestic computing power chips is insufficient. At the software level, the dominance of the mainstream deep learning framework ecosystem is still controlled by US enterprises, and domestic frameworks have gaps in compatibility with industrial-level complex tasks and developer ecosystem activity. At the toolchain level, AI tools dedicated to scientific computing highly depend on international open-source communities, and domestic AI for Science platforms have not yet formed unified standards and a complete ecosystem. In the field of core components, the localization rate of high-precision reducers, force sensors, and flexible actuators for humanoid robots is insufficient, and the low-temperature control systems, superconducting materials, and ion trap control components for quantum computing also seriously depend on imports. These shortcomings make it difficult for AI to deeply support original innovation breakthroughs in future industries.
2. Low efficiency of factor allocation, the value of new factors is not fully released. In terms of data elements, problems such as difficulty in rights confirmation, pricing, and circulation are prominent. Industry and regional barriers are serious. Data opening and sharing across industries, fields, and regions are insufficient. High-quality scientific research data and industrial data are scarce, making it difficult to fully release data value. In terms of computing power elements, the problem of unbalanced layout is prominent, showing a pattern of "strong in the East and weak in the West." The energy efficiency level of intelligent computing clusters is low, the supply of dedicated computing power for scientific computing is insufficient, and the degree of computing power universalization is difficult to meet the needs of large-scale simulation in future industries. In terms of talent elements, there is a serious shortage of high-end original innovation talents, interdisciplinary compound talents, and highly skilled talents. The talent cultivation system is disconnected from industrial needs, restricting the full release of innovation vitality.
3. Insufficient depth of industrial integration, low ecological collaboration efficiency. The integration of AI and future industries is still at a superficial stage. Most applications are limited to a single link and have not yet achieved intelligent transformation of the entire process and chain. Industrial chain collaboration is insufficient, with obvious chain breakpoints, link shortcomings, and collaboration barriers. The integrated development of large, medium, and small enterprises is restricted, and the leading and driving role of leading enterprises has not been fully exerted. The integration of "industry-university-research-application" is not deep enough. There is a disconnection between universities, research institutes, and enterprises. The transformation rate of scientific and technological achievements is low, and the problem of "valuing R&D while neglecting transformation" still exists. The construction of the open-source ecosystem is relatively lagging. There are few open-source platforms and tools with global influence, and the intensity of innovation resource integration is insufficient, making it difficult to form a synergistic innovation force.
4. Lagging construction of the standard system, insufficient international discourse power. The cross-integration characteristics of future industries lead to an imperfect standard system. The standard gap in the field of AI and cutting-edge technology integration is particularly prominent. There is a lack of unified standards in interoperability, security frameworks, identity authentication, and tool interfaces, restricting the large-scale application of technology and the coordinated development of the industry. The participation in international standards is low. China lacks discourse power in standards in fields such as quantum, biology, embodied AI, and AI for Science. It is difficult to lead the formulation of international rules, facing the unfavorable situation of "following standards and passive adaptation." The construction of safety, ethics, and governance standards in the field of AI lags behind, making it difficult to adapt to the needs of rapid technological development.
5. Insufficient adaptation of the governance system, weak risk prevention and control capabilities. Future industries are characterized by high risk, uncertainty, and cross-boundary nature. The traditional regulatory model by industry and field is difficult to adapt, and the problems of regulatory vacuum and regulatory overlap exist at the same time. Relevant ethical norms, laws and regulations, and regulatory rules are not yet perfect. Problems such as algorithm bias, data abuse, technology out of control, and increased security risks are becoming increasingly prominent. Risks such as supply chain security, data security, and network security brought by the rapid development of AI technology occur frequently. The construction of the risk prevention and control system lags behind, and the ability to prevent and resolve major risks needs to be improved.
03Building a New Ecosystem for Cultivating Future Industries Centered on AI
Based on the current situation of cultivating future industries in China and seizing the historical opportunity of AI, the industry urgently needs to focus on the four dimensions of technology foundation, factor reconstruction, ecological collaboration, and governance protection to build a systematic cultivation system covering the entire chain, all factors, and the whole ecosystem, driving China's future industries from following to running alongside, and from running alongside to leading, and accelerating the formation of new quality productive forces.
Technology Foundation: Strengthen the Leadership of AI for Science and Build a Solid Foundation of Core Technologies
1. Increase investment in basic research and strengthen the supply of original innovation. Give full play to the advantages of the new-type system for mobilizing resources nationwide. Focus on key fields such as quantum technology, biomanufacturing, embodied AI, brain-computer interfaces, and 6G. Continuously increase investment in basic research and steadily improve the proportion of basic research funding in the total R&D funding of the whole society. Layout and build a batch of national laboratories, frontier research institutes, and basic science centers, and concentrate efforts to break through a batch of principled and basic problems. Improve the scientific research evaluation mechanism, break the orientation of "only papers, only titles, only educational backgrounds, and only awards," establish an evaluation system centered on originality, breakthroughs, and practicality, and fully stimulate the innovation vitality of scientific researchers.
2. Focus on core technology research and break the "choke point" problems. Adhere to "industry asking questions and science answering questions." Focus on weakness fields such as high-end chips, core algorithms, key materials, precision components, and industrial software, and deeply implement new research mechanisms such as "open competition for the best candidates" and "horse racing system." Focus on breaking through the core technologies of AI, developing high-performance domestic large models, dedicated AI chips, AI for Science toolchains, and agent development platforms, and building an independent and controllable AI technology system. Relying on leading enterprises, specialized and sophisticated enterprises, and research institutes, form an innovation consortium with the deep integration of "industry-university-research-application," open up the innovation chain, improve the efficiency of technological research, and provide safe and reliable empowerment support for future industries.
3. Build AI for Science infrastructure and accelerate disruptive breakthroughs. Accelerate the layout of AI for Science platforms for future industries. Build industry-specific scientific large models around key fields such as quantum technology, biomanufacturing, new materials, and brain-computer interfaces, and drive AI to deeply integrate into the entire scientific research process. Support the construction of national scientific data platforms, intelligent experimental platforms, and digital twin scientific research platforms, and drive the formation of a new scientific research paradigm of "data-driven discovery and model-driven innovation." Build a batch of AI-empowered innovation platforms, concept verification centers, and pilot bases to accelerate the transition of disruptive technologies from the laboratory to industrialization.
Factor Reconstruction: Build a "Data-Model-Computing Power" Integrated Base and Release Factor Value
1. Deepen the market-oriented reform of data elements and release data value. Accelerate the market-oriented allocation reform of data elements, improve the institutional system of data property rights, circulation and trading, income distribution, and security governance, and break "data barriers." Build and improve national and industry-level data trading platforms, and drive the compliant circulation and efficient utilization of scientific research data and industrial data. Build a data security guarantee system, improve technical means such as data hierarchical and classified protection, privacy computing, and data desensitization, and balance data opening and security. Relying on AI technology, build an intelligent tool system for the entire chain of data collection, cleaning, labeling, and analysis, and improve the efficiency of data value mining.
2. Build an efficient and universal computing power network and consolidate computing power support. Coordinate and promote the "East Data, West Computing" project and the "West-to-East Power Transmission" strategy, accelerate the promotion of "computing-power and electricity coordination," optimize the spatial layout of computing power, and tilt computing power resources to the central and western regions to alleviate the imbalance of "strong in the East and weak in the West." Build a national integrated computing power network, coordinate the layout of intelligent computing clusters, supercomputing centers, and edge computing power nodes, focus on strengthening the construction of dedicated computing power for scientific computing, and realize on-demand scheduling, universal supply, green and low-carbon, and safe and controllable computing power. Increase the R&D of domestic AI chips and computing power servers, and build an independent and controllable computing power industrial chain. Lower the threshold for using computing power, launch universal computing power services, and support small and medium-sized enterprises and research institutions to obtain computing power resources at low cost.
3. Strengthen talent team building and cultivate compound talents. Deeply implement the projects of introducing top talents, cultivating young talents, training compound talents, and improving highly skilled talents, and build a multi-level and full-chain talent echelon. Focus on key fields of future industries, introduce a batch of top international scientists, leading talents, and innovation teams. Optimize the setting of disciplines and majors in universities, add interdisciplinary majors such as quantum technology, biomanufacturing, embodied AI, and AI, deepen school-enterprise collaborative education, and cultivate interdisciplinary compound talents. Improve the talent evaluation, incentive, and mobility mechanisms, increase the salary and treatment of scientific researchers, give scientific researchers greater decision-making power on technical routes and the right to use funds, and fully stimulate the innovation vitality of talents.
Ecological Collaboration: Build an Agent Innovation System and Improve Industrial Collaboration Efficiency
1. Build an intelligent collaborative innovation system for future industries. Adapt to the trend of AI evolving towards agents, drive the R&D of scientific research agents, industrial agents, design agents, manufacturing agents, and governance agents, and explore new models of human-machine collaborative innovation. Support the construction of agent development platforms and application ecosystems, and accelerate the formation of a new innovation system with the deep integration of "expert experience + industry knowledge + AI capabilities." Promote agents to become important production tools and innovation partners for future industrial innovation, and greatly improve innovation efficiency and boundaries.
2. Cultivate a high-quality enterprise echelon and strengthen the main role of enterprises. Make leading enterprises stronger and better, support chain-owner enterprises to integrate innovation resources, build AI innovation empowerment public platforms, lead industry standards, drive the industrial ecosystem, and give full play to the "leading goose effect." Support and expand specialized, refined, differential, and innovative enterprises, increase policy, funding, and technical support, encourage them to deeply cultivate segmented fields, overcome key technologies and key components, and form industrial resilience support. Vigorously cultivate technology-based small and medium-sized enterprises and innovative start-ups, lower the threshold for entrepreneurship, provide high-quality incubation services and financing support, and stimulate innovation vitality. Promote the integrated innovation of large, medium, and small enterprises, and build a collaborative development pattern of "chain-owner leading, specialized and sophisticated supporting, and start-ups active."
3. Build an open-source and open ecosystem and drive the sharing of innovation resources. Vigorously develop the open-source ecosystem, support enterprises, research institutes, and the vast number of developers to actively participate in open-source projects, and build a batch of open-source platforms, frameworks, and tools with global influence. Promote the integration of China's open-source technologies and products with the global ecosystem, and form an open-source development pattern of "open collaboration and mutual benefit and win-win." Relying on standardized protocols, build a technology interoperability ecosystem, break technological barriers, and drive the efficient flow of innovation resources across entities and fields. Encourage leading enterprises to open up technology, platforms, and data resources, and drive small and medium-sized enterprises to collaborate in innovation.
Governance Protection: Improve the Safe and Trustworthy Governance System and Ensure the Healthy Development of the Industry
1. Accelerate the construction of the standard system and enhance international discourse power. Adhere to the principles of standards first, independent innovation, open compatibility, and safe trustworthiness. Accelerate the construction of a full-chain standard system covering the technology, products, services, security, ethics, and governance of future industries, and focus on improving the standards in the field of AI and cutting-edge technology integration. Actively participate in the formulation of international standards, drive China's technology and product standards to become international standards, and enhance China's discourse power in the global governance of future industries. Relying on professional research institutions, take the lead in undertaking international standard projects, lead the formulation of international rules, and build a standard development pattern of "independent standard leadership and international standard participation."
2. Improve the security protection system and prevent various risks. Coordinate development and security, adhere to bottom-line thinking, and build a comprehensive and multi-level security protection system. Strengthen the security and controllability of core technologies, develop safe and reliable AI models and computing power chips, and prevent technology security risks. Improve the security protection mechanisms for data, network, and supply chains, strengthen data hierarchical and classified protection, vulnerability detection, risk warning, and emergency response capabilities, and prevent risks such as data leakage, network attacks, and supply chain interruptions. Establish AI security assessment, certification, and regulatory mechanisms, strengthen the governance of algorithm security, fairness, transparency, and interpretability, and prevent the risks of algorithm bias, abuse, and out of control.
3. Improve the regulatory governance system and realize inclusive and prudent regulation. Innovate regulatory models, break down regulatory barriers by industry and field, and build a collaborative regulatory mechanism across departments, fields, and regions. Adhere to "liberating innovation and holding the bottom line," implement inclusive and prudent regulation, "sandbox" regulation, and trigger-based regulation, and reserve sufficient space for technological innovation. Improve relevant laws and regulations, ethical norms, and industry self-discipline rules, clarify the boundaries of technology application and responsibility attribution, and prevent the risks of ethical misconduct and technology abuse. Strengthen international governance cooperation, actively participate in global future industry governance and AI governance, and drive joint consultation of rules, co-construction of standards, and joint promotion of industries.
04Conclusion
Future industries are the strategic engine for building a strong country and national rejuvenation, and the core support for cultivating new quality productive forces and building a modernized industrial system. Currently, the rapid iteration of AI technology provides unprecedented historical opportunities for the cultivation of China's future industries, and also brings systematic challenges. AI is increasingly becoming a key variable and core engine for the development of future industries. From scientific discovery to technological innovation, from resource allocation to industrial organization, and from value creation to governance system construction, AI is driving a systematic transformation of the cultivation logic of future industries. The cultivation of China's future industries has entered a critical window period. We must jump out of the traditional development paradigm, give full play to the advantages of China's super-large market, complete industrial system, and rich application scenarios, seize the opportunity of AI technological transformation, accelerate the construction of a new ecosystem for cultivating future industries led by AI for Science, driven by data intelligence, supported by an open ecosystem, and guaranteed by safe governance, drive China to achieve the leap from following and running alongside to leading in the global competition of future industries, and provide strong momentum for building a modernized industrial system and achieving high-quality development.
This article is published in the 7th issue of "Communications World" in 2026; Original title: "Building a New Ecosystem for Cultivating Future Industries Empowered by AI"; Authors: Liu Tengfei, Liu Tongtong, Liu Naijia, Future Industry Department, Institute of Industrial and Internet, China Academy of Information and Communications Technology.
Editor-in-Charge/Layout: Zhu Wenfeng
Proofreader: Mei Yaxin
Supervisor: Liu Qicheng