Artificial Intelligence (AI) has become a vital determinant of economic competitiveness and state capacity in the digital age. In this context, AI infrastructure through a combination of data systems, computing infrastructure, policy frameworks, regulations, and institutional mechanisms, enable the development and deployment of AI at scale. Countries that invest in these foundational layers are better positioned to lead in innovation, industrial upgrading, and governance in an increasingly digital global economy. China presents an example of how government-driven AI infrastructure investments and policies have systematically driven AI advancement and global competitiveness. Through coordinated planning, long-term policy direction, and consistent public investment Beijing has laid the foundation of its AI industry. An integrated ecosystem promotes research, data circulation, computing capability, and practical applications across industries in China by coordinating national policies with industry priorities.
Data Marketplace
In the Opinions on Building a More Complete System and Mechanism for Market-Based Allocation of Factors released in April 2020, China categorised data factors as the fifth production component after land, labour, capital, and technology. Such an outlook is linked to aspects of the digital economy, where data is increasingly commodified and used for financial services, retail sales, business development, and digital innovation. In fact, the Action Plan for Promoting Development of Big Data launched in 2015, was the first national policy document that articulated the potential of data as a major production resource. Following the Action Plan, China started promoting big data projects in different industries with new data-driven formats, like internet finance, services, pharmaceuticals and data processing. For instance, in the education sector, projects included establishing electronic student records and national databases and building digital facilities for libraries, archives and museums. Initiatives in the healthcare industry involved building electronic health records and medical databases. Provinces like Guizhou launched big data center exchange pilot programs; however, they remained largely ineffective as most transactions occurred informally over-the-counter. Moreover, due to significant compliance risks and high operating costs, the absence of uniform legal and policy norms among local data exchanges reduced business incentives. Without clear legislation defining data rights and ownership, companies faced legal uncertainty and thus discontinued data procurement due to concerns about illegal over-collection and lack of compliance guarantees.
Drawing from the lessons of the Action Plan, in September 2021, the Data Security Law (DSL) was released, reaffirming state-backed establishment of data center exchange. Under the DSL, personally identifiable information (PII), sensitive information on national security and public interest, were restricted and not allowed to be traded with domestic or foreign entities. Moreover, the 14th Five Year Plan (2021-2025), expanded sectoral big data platforms and applications by encouraging data sharing, innovation competitions, pilot projects, and training programs. It also supported the big data ecosystem through capability standards, open-source communities, talent development, and monitoring systems to improve industry oversight and data service development.
Data exchange centers created a standardized and transparent environment for data circulation in China, by acting as an intermediary marketplace that verifies ownership rights, ensures secure transactions, and sets rules for data usage. Major cities developed their own data exchange platforms while following strict guidelines. A notable example is the Shanghai Data Exchange (launched in 2021 for businesses), with transaction data worth 1.16 billion Yuan in 2023. It functions on a Data Capital Bridge System - “one bridge, two locations, two axes”, with the framework of 'Data Easy Loan' (数易贷) and 'Data Asset Connection' (数资通). It structures raw data into processed datasets which can be used for specific cases, thus making data products. Furthermore, to make these data products traceable the system issues registration certifications to confirm ownership, creating a "data alliance chain". Beijing has promoted multiple such regional data exchanges and new regulatory bodies to support regulated data trading and cross-industry sharing, reflecting a national strategy to build a structured and state-guided data marketplace.
Data Centres and Compute Capacity
China's data center expansion combines energy efficiency with regional development. Currently, China hosts around 561 data centers, run by 34 operators, with the majority of them clustered in the eastern coastal regions of Guangdong, Shanghai and Beijing. These regions have a large volume of digital activity due to dense populations, modern industry, and are home to leading internet companies. However, expanding large-scale data infrastructure further is becoming more challenging in China’s eastern provinces due to increased electricity costs, land availability, and environmental challenges.
In order to address this challenge, the government launched the Eastern Data Western Computing strategy (东数西算) in 2022, which shifts a portion of data processing and storage to western provinces while keeping demand centers in the east. The policy established eight national computing hubs and ten major data center clusters, including hubs in Inner Mongolia, Guizhou, Ningxia, Gansu, and Sichuan, with investments of over 43.5 Billion Yuan. Western regions offer abundant land, lower temperatures, and access to renewable energy, thus reducing cooling and operational costs for large data centers. This geographic redistribution of data centres also aligns with broader development goals, by directing digital infrastructure investment into less-developed provinces, thus reducing regional economic disparities while supporting China’s rapidly growing cloud and AI industries.
AI in Education
China has been actively advancing AI development since 2017, recognizing its importance to industry and trade early on. The launch of the New Generation Artificial Intelligence Development Plan in July 2017 outlined an agenda to integrate AI as a formal discipline across all levels of education, from primary schools to universities, building a future-ready talent pipeline. China is rapidly integrating AI through a centralized approach that spans classroom instruction, teacher empowerment, and administrative resource allocation; with policies such as Smart Education of China 2021 initiative, and National Education Master Plan 2035. In practice, students are utilizing AI music software in Xinghua middle schools, intelligent sports hardware for running practice in Qinhuangdao primary schools, and targeted cognitive interventions like the SMART program to enhance reasoning and decision-making skills in high schools. Simultaneously, AI is drastically reducing educators' workloads; physics teachers at Xiamen Shuangshi High School use AI to quickly generate dynamic simulation code, and Gulin Township Junior High School employs an AI grading machine to instantly flag common errors and analyze individual performance, supported by a Ministry of Education initiative that trained over 2.97 million teachers by the end of 2024 and developed over 700 AI tools. Beyond the classroom, local governments in Baotou and Hangzhou's Fuyang District utilize AI-driven data analytics to evaluate teacher demographics, identify staffing gaps, and equitably allocate educators.
AI Research Ecosystem
China is strengthening its AI infrastructure through state-supported research labs, university-industry partnerships, and large-scale automation across sectors. Universities are collaborating with big technology companies like Baidu, Huawei, Alibaba, and Tencent to integrate AI research into academic programs, while creating industry-led innovation and real-world applications. A notable example is the partnership between Tsinghua University and Huawei, which focuses on developing AI chips optimized for deep learning, supported by funding, advanced technology access, and knowledge exchange. Similarly, Zhejiang University’s collaboration with Alibaba applies AI in healthcare to improve diagnostics, patient care, and hospital management.
China now holds 60% of the world’s AI patents. In 2025 alone, it had over 6,000 AI businesses, with the industry estimated at more than 1.2 trillion yuan. To support governance and responsible development, the government launched the Beijing Algorithm Registration Center (北京算法登记服务中心) in July 2024 to provide a physical platform for the registration of algorithms and large language models (LLMs) while offering technical assistance, intellectual property guidance, and support related to data trading. Regionally, China fosters AI research through targeted incubator hubs that are increasingly supporting the booming trend of AI-powered "one-person companies" (OPCs). For example, Beijing's ZGC AI North Latitude Hub helps startups and OPCs with company registration, legal consultation, tech-matching, and AI training. Similar hubs in Hangzhou also nurture OPCs, while Shenzhen pioneers brain-science technology to shape future industries. Fueled by these regional ecosystems and generative AI tools, over 16 million OPCs now operate nationwide.
Policy and Strategic Direction
China’s approach to artificial intelligence has developed progressively since 2017, with the state playing a central role in guiding both infrastructure and regulation. The New Generation Artificial Intelligence Development Plan of 2017 laid the foundation by setting the goal of global leadership by 2030. This was reinforced by the 2019 Guidance on National New Generation Artificial Intelligence Open Innovation Platform Construction Work issued by the Ministry of Science and Technology which aims to drive AI technical innovation and industrial development by empowering leading tech companies to build open-source AI platforms. These platforms propose to integrate resources, provide open services, and lower the barrier of entry for small businesses and entrepreneurs. From 2020 onwards, China shifted towards building governance frameworks alongside AI development. The Guidelines for Establishing a National AI Standard System marked the beginning of structured standard-setting. This was followed by the New Generation AI Ethics Specifications (2021), the first specification on AI ethics to ensure it is always under human control and trustworthy, following a series of data and cybersecurity regulations.
By 2024, the AI Safety Governance Framework (V1.0) further strengthened oversight. In 2025, new national standards such as GB45438 for AI-generated content labelling and GB/T 45652 for security on pre-training and fine tuning data on generative AI models were announced. The “Artificial Intelligence+” initiative represents the latest phase in China’s AI development policy plans, with a focus on encouraging enterprises to adopt AI as a core operational tool rather than a supplementary technology. This reflects a broader transition from earlier phases that focused on building foundational tools through research institutions and national labs, to one centered on practical deployment across industries. This is particularly applicable in manufacturing, to improve productivity and cut costs through automation, quality control, and predictive maintenance. AI is also being used by service industries like banking, healthcare, retail, and logistics for supply chain optimization, risk assessment, consumer interaction, and diagnostics.
In addition to a top-down approach, local governments are pushing for AI applications with schemes like Compute Vouchers (算力券), proposed by Beijing, Shanghai and Shenzhen. They offer vouchers typically worth between $140,000 to $200,000, which allows companies to redeem time in data centers to run and train their new AI models, thus significantly reducing their costs. Similarly, provinces are also providing incentives to train youth and have an AI capable workforce. For instance, the new talent policy in Shanghai offers welcome packages, free accommodation spaces and rent-free areas for startups for up to three years. Similarly, Zhuhai aims to target fresh graduates by giving them free housing for the first year and a 70 percent rent reduction in the second year.
Leading AI Companies
Public and private enterprises have also been key facilitators of China’s AI infrastructure, especially with the emergence of national champions closely aligned with state priorities. By creating domestic AI chips (Ascend series), computing hardware, and full-stack AI software, companies like Huawei play a key role in creating a domestic AI ecosystem and lowering dependency on foreign technology. This push is evident in large-scale deployments of Ascend chips across data centres and AI training systems.
China Telecom, which supports the network backbone and data center ecosystem at the infrastructure level, recently demonstrated that it can train large-scale AI models solely on local chips, highlighting China's transition to independent AI systems. Meanwhile, Alibaba Group leads in cloud computing and AI platforms, building full-stack capabilities spanning foundation models, cloud infrastructure, and proprietary chips to support enterprise and consumer AI applications. They have emerged as a key player with its Qwen models, which are widely deployed across cloud services and e-commerce and that operate directly as intelligent smartphone assistants. Their commercial impact is evident, with AI systems processing over 120 million orders within six days during peak retail activity, demonstrating strong real-world integration. Tencent, meanwhile, is leveraging its vast user ecosystem through WeChat to scale its Hunyuan AI models and newer AI agent systems. It plans to distribute and integrate AI directly into messaging, payments, and enterprise tools, reaching over a billion users.
Baidu has engaged with the competition by making its Ernie Bot open-source, embedding AI into search, autonomous driving, and cloud, combining data advantage with long-term AI research. At the same time, newer players like DeepSeek are intensifying competition by building high-performance, lower-cost models, challenging incumbents on efficiency. This landscape can be understood as a race between major firms to build foundation models and quickly embed them into large user ecosystems. Unlike a pure research race, competition is driven by who can commercialize AI fastest and at scale.
In the colocation data center market, companies like Zenlayer and China Telecom dominate, while Equinix and Amazon Web Services follow closely. The competition among these companies demonstrates a coordinated paradigm in which state policy, infrastructure, and industry innovation come together to grow AI capabilities across China.
What Next: 15th Five Year Plan
China's 15th Five-Year Plan emphasises that AI and digital infrastructure are key drivers of economic development. It focuses on new sectors such as robots, embodied AI, biomedicine, and the low-altitude economy, and pledges to increase investment in fundamental research, through greater percentage of total R&D expenditure and the establishment of new large-scale scientific institutions. The plan broadens the integration of AI across industrial production, services, and government, building on previous projects like "Artificial Intelligence+." It reflects a shift in approach from viewing technology transformation as sectoral supports, to integrating them as fundamental inputs for production. Expanded data infrastructure, enhanced digital ecosystems, and unified national computer networks have been given top priority. The 15th Five-Year Plan integrates AI and digital systems more thoroughly into the foundation of China's industrial and innovation framework, in contrast to the 14th Five-Year Plan which focused on smart manufacturing.
At the same time, emerging markets such as quantum computing, brain-computer interfaces (BCI), embodied AI, and 6G are being developed as future growth engines, with entrepreneurs estimated to generate more than 10 trillion yuan by 2030. Innovation hubs like Shanghai are fostering innovation by partnering hospitals with technology companies to accelerate real-world applications, especially across advanced healthcare sectors. Major tech firms like Xiaomi and XPeng are championing embodied AI by designing humanoid robots to work directly on their own assembly lines. To further accelerate this commercialization, regulators are employing "sandbox regulation" to safely test innovations and are actively relaxing restrictions to open new markets for technologies like robotaxis, city delivery drones, and private rockets.
Conclusion
China’s AI ecosystem is best understood as a both state-coordinated and privately executed system, where infrastructure expansion is rapid, targeted, and closely aligned with national priorities. This model has clear strengths: strong policy continuity, large-scale state funding, and the ability to mobilize firms like Huawei and Alibaba to build chips, cloud, and data infrastructure in parallel. This ecosystem enables rapid deployment of AI across sectors, from manufacturing to governance.
At the same time, despite significant progress, China faces specific and pressing challenges. The AI ecosystem is highly resource-intensive, and its rapid development has increased pressure on energy and environmental resources, with data centres requiring significant electricity and water for cooling, contributing to carbon emissions. This creates a clear tension between digital growth and China’s broader climate commitments. Additionally, dependence on advanced semiconductors and external technologies continues to limit full self-reliance. These challenges are compounded by a dense and evolving regulatory environment which supervises the operation of algorithms, and data security; which can slow experimentation and innovation. Taken together, China’s experience highlights the value of integrated infrastructure planning, strong public-private coordination, and early investment in core technologies. At the same time, it highlights the need to balance scale with sustainability, ensure more equitable regional development, and maintain regulatory flexibility in a rapidly evolving global technology landscape.
Image Credits: Xinhua & China Daily
Author
Stuti Agarwal
Stuti Agarwal is a Research Intern at the Organisation for Research on China and Asia (ORCA). She is a final-year Business Analytics student with a minor in Public Policy at FLAME University. Her work involves supporting research through data collection, analysis, and visualization. She has previously worked with labour and wage datasets in a market research setting. Stuti’s academic interests include environmental policy, sustainability, and the role of data in policy research. She is interested in using clear and accessible visuals to aid research communication and interpretation.