šØš³ Chinaās National AI Strategy
Chinaās AI push has shifted from catching up at home to setting the rules abroadāmelding chip self-reliance, state capital, and Digital-Silk-Road diplomacy into a single strategic engine.
Chinaās AI strategy is built on achieving global leadership by 2030, driven by high-level political support, deep state-led investment, and foundational policies like the 2017 New Generation AI Development Plan and Made in China 2025.
The strategy prioritizes self-sufficiency in critical technologies, especially semiconductors and AI chips, in response to geopolitical constraints, while promoting homegrown AI models like DeepSeek and expanding AI integration across sectors.
China is exporting its AI governance model and infrastructure globally, particularly through Digital Silk Road initiatives and AI capacity-building programs in the Global South, aiming to shape international norms and standards.
What it means for policy-makers and investors
- Regulatory first-mover advantage. China is legislating faster than the West, giving its platforms a playbook that could gain extraterritorial clout as they expand abroad.
- Compute bifurcation. Washingtonās export controls have hardened Beijingās resolve to own the hardware stackāfrom EDA tools to data-centre clusters in Gansu. Expect heavy capex and selective over-capacity risk.
- South-South alignment. Training programmes and Digital Silk Road tie-ups are embedding Chinese technical standards in emerging markets, shaping everything from data-governance norms to procurement preferences.
Bottom line: Chinaās AI agenda has morphed from factory-floor automation to a whole-of-nation contest for algorithmic influence. Silicon sovereignty remains its Achilles heel, but the country is now as much a rule-setter as a fast follower.
Strategy Evolution
What began as an import-substitution project is now a bid to write the rule-book. Beijing has paired lavish industrial policy with ever-tighter regulation and an active āAI diplomacyā push in the Global South. The timeline below distils the main inflection points.
2025. Global outreach & strategic self-sufficiency
DeepSeek and a dozen other home-grown LLMs are rolled out at scale; ten Cyberspace Administration workshops target developing nations; U.S. chip curbs keep pressure on domestic fabs.
2024. Capacity-building & norm-setting
AI Capacity-Building Action Plan for Good and for All puts the UN centre-stage while Huaweiās Ascend 910C courts buyers shut out of Nvidiaās H100.
2023. Generative-AI rule-book
The Interim Measures for Generative-AI Services impose data-source disclosure, watermarking and security assessmentsāmonths before Brussels finalised the EU AI Act.
2022. From algorithms to compute topology
Two moves stand out: (i) the Deep-Synthesis Provisionsāthe worldās first mandatory deepfake labelling regime; (ii) East DataāWest Compute, an eight-hub programme that shifts workloads inland to cheaper, greener power.
2021. Algorithmic governance & FYP lock-in
The Algorithm-Recommendation Regulations force platforms to disclose ranking logic; the 14th Five-Year Plan elevates AI to ābasic infrastructureā.
2020. Pandemic test-bed & ānew infrastructureā stimulus
Covid-19 accelerates computer-vision triage and QR health codes; Beijing earmarks >Ā„1 tn for 5G, cloud and AI under the ānew infrastructureā banner.
2019. Race of two giants
Officials declare a mano-a-mano contest with America, yet fret about elite talent and 7-nm chips.
2018. Institutional build-out & military integration
NUDT opens twin AI labs; the CAICT warns of an āintelligentisedā arms race.
2017. New-Generation AI Development Plan
Targets world-leading status by 2030 and ties civilian R&D to āleap-frogā defence gains.
2015. Made in China 2025 & the first chip fund
AI and semiconductors become cornerstones of tech independence.
Pre-2015. Vision & early bets
Xi Jinping hails AI as a ānew engine of rejuvenationā; state funds seed patent sprint.
2025 (January to June)
* Ongoing focus on challenges posed by U.S.-led semiconductor restrictions.
* Homegrown AI models like DeepSeek continue to be positioned as alternatives to Western models, emphasising cost-effectiveness and scalability.
* Continued integration of AI across various sectors including surveillance, healthcare, and smart cities.
* Assessment of China's AI Strategy provides key insights into ongoing trends.
* Ten AI workshops and seminars primarily aimed at developing countries are planned to be held by the end of 2025.
As of early to mid-2025, China's AI strategy continues to be heavily influenced by the foundational documents, the New Generation AI Development Plan (AIDP) from 2017 and the Made in China 2025 initiative. A significant aspect remains the ambition to position China as a global leader in AI innovation. Homegrown models such as DeepSeek are highlighted as providing cost-effective and scalable alternatives to Western counterparts like OpenAI's ChatGPT. DeepSeek focuses on cost-effective operations through leaner training methodologies and smart architectural designs, challenging traditional high-cost approaches in AI development. This push for indigenous models is part of the broader goal of technological self-sufficiency.
The integration of AI technologies continues across various sectors, including surveillance, healthcare, and smart cities. In healthcare, AI is applied through mobile health technologies for tailored therapy, wearable interfaces for monitoring, medical imaging interpretation, and predictive modeling. However, challenges related to data protection and interoperability remain important considerations for AI integration in healthcare.
A major challenge persists due to U.S.-led semiconductor restrictions, which pose significant hurdles for China's ambitions by limiting access to cutting-edge technology crucial for advanced semiconductor capabilities. These restrictions aim to hinder China's progress in semiconductor design and manufacturing.
Globally, China is actively using initiatives like the Digital Silk Road (DSR) to export its AI technologies and governance models. The DSR, intersecting with Made in China 2025 goals, aims to enhance global supply chains through AI and Big Data integration, serving both commercial and potential military objectives. This effort is seen as reshaping AI standards and practices in developing economies, amplifying China's influence. China's AI governance model, balancing innovation and social stability, is attractive to developing nations, and its dual-track regulation strategy involves both international cooperation and promoting its own standards through bilateral projects. This could lead to alignment of AI standards with China's in these regions, bringing economic growth but requiring careful monitoring of privacy and civil liberties.
China's Provisional Administrative Measures of Generative AI Services, being the first regulation of its kind, are viewed as potentially setting a precedent for international AI policy discussions. While aiming for a unified approach, these measures also risk creating ambiguities in accountability across the AI value chain.
The Chinese government continues to focus on capacity-building efforts for developing countries, with a commitment to hold **10 AI workshops and seminars primarily aimed at these nations by the end of 2025**.
2024
* The **AI Capacity-Building Action Plan for Good and for All** was announced, calling for increased investment in AI capacity-building globally, with a focus on the Global South and the UN's central role.
* This plan outlines specific visions and China's proposed actions for international AI cooperation across infrastructure, applications, talent development, data, and safety.
* China leads the world in generative AI patents.
* Investments in Chinese AI companies remain substantial.
* Notable national AI teams and new startups continue to drive dynamism in the private sector.
* Huawei's Ascend 910C chip is positioned as a direct competitor to Nvidia's H100.
* Shift in consumer sentiment and growing optimism regarding China's economic future, partly linked to perceived success in overcoming sanctions and innovation in areas like AI and robotics.
A significant development in 2024 was the announcement of the **AI Capacity-Building Action Plan for Good and for All**. This plan, updated on September 27, 2024, highlights China's belief in the importance of upholding the United Nations' central role in international development cooperation and pursuing true multilateralism to bridge the AI and digital divides, particularly for the Global South. It promotes implementing relevant UN resolutions and the 2030 Agenda for Sustainable Development through various forms of cooperation, based on principles like sovereign equality, development orientation, shared benefits, and inclusiveness.
The Action Plan outlines a vision including promoting AI and digital infrastructure connectivity, especially assisting the Global South; empowering industries through "AI Plus" applications in manufacturing, agriculture, green transition, climate change response, and biodiversity conservation; enhancing AI literacy and strengthening personnel training through education, professional exchange, public literacy campaigns, and protecting the digital rights of women and children; improving AI data security and diversity by promoting orderly data flow, exploring global data-sharing platforms, protecting privacy, and eliminating bias in datasets; and ensuring AI safety, reliability, and controllability through global risk assessment frameworks, standards, and governance systems under the UN framework, taking into account developing countries' interests.
China's proposed actions in the plan include engaging in AI capacity-building cooperation, jointly implementing UN Summit of the Future outcomes, and building joint AI infrastructure and laboratories, particularly with developing countries. China is also ready to cooperate on AI R&D and applications in areas like poverty reduction, healthcare, agriculture, and education, aiming to unlock the dividends of AI as a "new quality productive force". Other actions involve tapping AI potential in green development and climate change response, establishing an international platform for AI capacity-building, sharing best practices, building open-source communities, holding AI capacity-building programs and workshops for developing countries, promoting public AI literacy with a focus on women and children, developing AI language and data resources while combating discrimination, and strengthening policy exchanges on AI strategies, testing, evaluation, certification, and regulation to address risks.
In terms of global AI development, China is positioned as a formidable competitor to the United States. **China leads the world in generative AI patents**, having filed over 38,210 patents from 2014 to 2023, significantly more than the United States. Despite having fewer AI companies than the U.S., Chinese firms have received substantial investments, totalling around $85.65 billion from 2014 to 2024. Notable private sector players like Baidu, Tencent, Alibaba, SenseTime, and iFlytek have been designated "national AI teams," tasked with leading advancements in specific sectors. New startups such as Zhipu AI and Baichuan AI also contribute to the dynamic ecosystem.
Geopolitical tensions, particularly U.S. sanctions, continue to influence China's AI chip market, driving a strategic shift towards self-sufficiency. While China's semiconductor manufacturing lags global leaders by several generations, companies like Huawei are developing alternatives to restricted foreign products. Huawei's Ascend 910C chip is reported to be competitive with or superior to Nvidia's H100 for AI workloads, marking a significant step in reducing reliance on imports.
Beyond official strategies, observations suggest a shift in consumer sentiment and growing optimism in China in 2024, with people starting to regain confidence and increase spending after the COVID era. This is linked to perceptions of overcoming U.S. sanctions and seeing China become a "locus for innovation" in AI and robotics. The adoption of robots in factories and their visibility in everyday life are cited as examples of this innovation. Analysts estimate China will contribute a third of global growth over the next five years.
Top AI tools in China include Baiduās Ernie Bot (a ChatGPT equivalent), models from Tsinghua University-affiliated startups like Zhipu AI and Baichuan AI, Alibabaās DAMO Academy for e-commerce AI, Huaweiās Cloud AI Solutions for enterprises, SenseTime and Megvii (Face++) for computer vision and facial recognition, iFlytek for speech recognition, and AI integrated into platforms like Zhihu.
2019
* Analysis from early 2019 highlighted China's significant progress in AI R&D and commercial applications, viewing itself in a "race of two giants" with the US.
* China's leadership was seen as viewing AI as a potential military "leapfrog development" opportunity.
* Weaknesses were identified in top talent, technical standards, software platforms, and semiconductors.
* Government efforts to improve the AI talent pool were underway.
* The importance of AI chips for future AI competition was recognised as strategic.
* Adverse macroeconomic factors and a potential financial bubble were noted as potential slowdowns for the AI sector.
In early 2019, analysis of China's AI strategy indicated that Chinese leadership viewed being at the forefront of AI technology as critical for future global military and economic power competition. The core strategy was still underpinned by the AIDP (2017) and Made in China 2025 (2015). China had significantly increased national and local government spending on AI, with regional governments committing billions. Chinese policymakers demonstrated a keen awareness of AI industries and policies in other countries, particularly the United States, translating and analysing U.S. reports.
China's goal was clear: to pursue global leadership in AI and reduce dependence on foreign technology imports. President Xi Jinping emphasised the need for China to be in the "front ranks" of theoretical research and grasp critical and core AI technologies.
While expressing concern about AI arms race dynamics in diplomatic forums, Chinese officials and government reports also saw increased military usage of AI as inevitable and were aggressively pursuing it. China was already exporting armed autonomous platforms and surveillance AI. The military aimed to narrow the gap with global advanced powers using "intelligent technology". Defense industry executives predicted future battlefields with no human fighters and lethal autonomous weapons being commonplace by 2025. AI was also being explored for command decision making, potentially leading to an "AI cluster" dominating command structures. Extensive use of AI in domestic surveillance, notably in Xinjiang, was widely acknowledged.
China's government and industry believed they had largely closed the gap with the United States in AI R&D and commercial applications, seeing themselves in a "**race of two giants**". By mid-2018, China's leadership assessed the country had entered the "first echelon" internationally in AI industry competitiveness. Studies indicated China led in total AI research papers, highly cited papers, AI patents, and AI venture capital investment, ranking second in the number of AI companies and AI talent pool size. Commercial successes in areas like computer vision (SenseTime) and consumer drones (DJI) demonstrated innovative products competitive on the global market. These commercial successes were directly relevant to national security, military, and espionage capabilities, as Chinese firms cooperate extensively with state security services under initiatives like Military-Civil Integration.
However, key weaknesses relative to the U.S. were recognised, particularly in **top talent, technical standards, software platforms, and semiconductors**. While China had a large AI talent pool, it lagged significantly in top-tier talent. The government was actively addressing this through initiatives like the Ministry of Education's **AI Innovation Action Plan for Colleges and Universities**, launched in April 2018, which aimed to create new teaching materials, online courses, and AI research centers. A five-year program to train more AI instructors and top students was also planned.
China's AI ecosystem relied on international markets, technology, and research collaboration. Many achievements were multinational, and Chinese researchers frequently coauthored with non-Chinese individuals. Chinese companies, even market leaders like DJI, had significant dependencies on U.S. components, particularly semiconductors.
The lack of strong domestic capabilities in software frameworks (like TensorFlow or PyTorch) was seen as a weakness, with domestic R&D often relying on foreign platforms. In semiconductors, despite being central to electronics manufacturing globally, China was heavily dependent on foreign chips designed elsewhere and manufactured in Taiwan or Korea. While Chinese companies were making progress in semiconductor design (like Huawei's HiSilicon), manufacturing capabilities lagged global leaders by several years.
The importance of **semiconductors, especially custom AI chips**, was increasingly recognised as critical for the future of AI competition. AI performance, particularly for training and synthetic data generation, was limited by computing power, making access to high-performance computing systems and advanced chips crucial. While China was strong in high-performance computing integration, it lacked major domestic manufacturers of advanced GPUs. However, the rise of custom AI accelerator chips offered an opportunity, as these could provide superior performance using older manufacturing processes, aligning with China's existing capabilities. Chinese tech giants and startups were investing in AI chip design divisions. Despite being behind in semiconductors, the trend suggested the gap would narrow due to government priority and massive investment, including dedicated national investment funds.
Potential challenges included adverse macroeconomic factors and concerns about a **financial bubble** in the AI startup sector, where high valuations might not be sustainable based on profitability, potentially slowing future R&D investment.
2018
**Key Developments:**
* Multiple high-level events and reports focused on AI, including a Politburo study session led by Xi Jinping.
* Ministry of Education launched the **AI Innovation Action Plan for Colleges and Universities**.
* Ministry of National Defense established two major AI research organizations under NUDT.
* CAICT published the **Artificial Intelligence Security White Paper**, calling to avoid AI arms races but also highlighting reliance on foreign software frameworks.
* China's position paper on lethal autonomous weapons was submitted to the UN.
* Alibaba cofounder Jack Ma publicly noted the importance of core technologies like chips due to foreign control.
2018 was a pivotal year with significant activity and assessment of China's AI landscape. In October 2018, President Xi Jinping led a Politburo study session specifically on AI, reiterating the goals of world leadership and self-reliance outlined in the AIDP and Made in China 2025. This demonstrated the high priority AI held for China's leadership.
April 2018 saw the launch of the **Ministry of Education's AI Innovation Action Plan for Colleges and Universities**, a key government initiative aimed at addressing the perceived weakness in top AI talent by significantly enhancing China's university AI curricula and training programs.
In the military sphere, the Ministry of National Defense established two major research organisations focused on AI and unmanned systems ā the Unmanned Systems Research Center (USRC) and the Artificial Intelligence Research Center (AIRC) ā under the National University of Defense Technology (NUDT). These centres rapidly grew to over 100 staff each, becoming among the largest and fastest-growing government AI research organisations globally. Military leaders increasingly adopted the term "intelligentized" warfare, reflecting the expected future basis of conflict driven by AI.
Concerns about AI arms races were voiced in diplomatic forums. Fu Ying, Vice-Chair of the Foreign Affairs Committee of the National People's Congress, spoke of the "threat of the new [AI] technology to mankind" and the need for cooperation. Chinese officials worried about AI lowering the threshold for military action and increasing the risk of misperceptions. The China Academy of Information and Communications Technology (CAICT) published its **Artificial Intelligence Security White Paper** in September 2018, which included a call to "avoid Artificial Intelligence arms races among countries". However, China's actions, including the development and export of armed autonomous systems, appeared to contradict these concerns. China's April 2018 position paper submitted to the UN on lethal autonomous weapons supported a ban but used a definition so narrow it seemed designed to allow China's own development to continue.
Chinese companies were increasingly recognised for their innovative AI products and services. SenseTime, a national champion in computer vision, reported massive revenue growth and was a major provider and exporter of surveillance technology. The dependence of Chinese AI development on foreign technology, particularly semiconductors and software frameworks (like those from Google and Microsoft), was a frequently discussed weakness. This dependence was underscored by the ZTE export restrictions imposed by the U.S.. Alibaba cofounder Jack Ma publicly stressed the critical need for China to develop its own core technologies to avoid such vulnerabilities.
Discussions in 2018 also highlighted the growing strategic importance of AI chips, with Chinese firms like Baidu, Alibaba, and Huawei establishing dedicated design divisions. The government's commitment to reducing dependence on foreign semiconductors was evident in the launch of a second national integrated circuit industry investment fund worth 300 billion RMB.
Towards the end of the year, reports of layoffs in China's tech sector and falling office real estate prices in tech hubs suggested potential macroeconomic headwinds and concerns about a financial bubble in the venture capital-fueled AI startup ecosystem.
2017
* The **New Generation Artificial Intelligence Development Plan (AIDP)** was issued by the State Council in July, becoming a core document of China's AI strategy.
* The AIDP set ambitious goals for AI industry competitiveness and technological development.
* The AIDP explicitly linked AI development to national security, economic competitiveness, and leapfrog development opportunities.
* The AIDP called for international cooperation on AI laws and regulations.
* AlphaGo's victory over Lee Sedol in 2016 was publicly cited by the Chinese military as demonstrating AI's potential for command decisionmaking.
July 2017 marked a pivotal moment with the release of the **New Generation Artificial Intelligence Development Plan (AIDP)** by China's State Council. This document, alongside the earlier Made in China 2025 initiative, formed the foundation of China's comprehensive AI strategy. The AIDP explicitly stated that AI had become a "**new focus of international competition**" and a "strategic technology that will lead in the future," necessary to enhance national competitiveness and protect national security.
The AIDP set ambitious goals, aiming for China's AI industry competitiveness to enter the "first echelon internationally" by 2020. It called for achieving world-leading levels in AI technology and reducing vulnerable dependence on imports. The plan emphasised AI as a "**major historic opportunity**" for national security "leapfrog development," suggesting it could offer military advantages over the U.S. and be easier to implement in China.
The AIDP also addressed the international dimension, stating that China would "deepen international cooperation on AI laws and regulations, international rules and so on, and jointly cope with global challenges".
While published in 2017, the AIDP and Chinese military thinking reflected insights from events like AlphaGo's victory over Lee Sedol in March 2016. A publication by the Central Military Commission Joint Operations Command Center highlighted AlphaGo's win as demonstrating the "enormous potential of artificial intelligence in combat command, program deduction, and decisionmaking".
2015
* The **Made in China 2025** initiative was released in May, becoming a core document outlining China's strategy for industrial upgrading and technological independence, with AI as a key component.
* The initiative set goals for reducing dependence on foreign technology.
* The first national integrated circuit industry investment fund was established to boost domestic semiconductor manufacturing.
The **Made in China 2025** initiative, released in May 2015, predated the AIDP but served as a foundational document that heavily influenced China's AI strategy. It outlined China's plans to upgrade its manufacturing capabilities through the integration of information technology, aiming to enhance productivity and increase the indigenous content in higher-end technology products. AI was central to this initiative, which sought to achieve technological independence and **diminish reliance on foreign inputs**. The goal was to secure a dominant position in the global tech landscape and position China as a science and technology superpower by 2049.
Also in 2015, China's government established the **first national integrated circuit industry investment fund**, a significant financial commitment (138.7 billion RMB or ~$20.5 billion) aimed specifically at reducing China's dependence on foreign semiconductors and boosting domestic manufacturing capabilities. This initiative highlighted the early recognition of semiconductors as a critical bottleneck and a strategic priority for technological self-sufficiency.
Prior to 2015
While specific policy documents are detailed from 2015 onwards, China's strategic thinking on AI was developing earlier. According to analysis from 2019, Chinese leadership already believed being at the forefront in AI technology was critical for the future of global power competition. The issue of AI had received significant attention from the highest levels of leadership, including Xi Jinping, even before the formal release of the AIDP. The patent data mentioned in Source, showing China's lead in generative AI patents, begins in 2014, indicating activity and investment in key AI areas predating the formal strategies.
In summary, China's AI strategy has evolved significantly since 2015, underpinned by ambitious national plans. Starting with foundational goals of self-sufficiency and industrial upgrading (Made in China 2025) and accelerating into a comprehensive national strategy encompassing R&D, industrialisation, talent, security, and international cooperation (AIDP), China rapidly positioned itself as a major global player. More recent developments highlight a continued drive for leadership, particularly in critical areas like AI chips, while simultaneously engaging in international capacity-building efforts, particularly with the Global South, and navigating complex geopolitical challenges.
