Abstract
Leading military forces around the world have expressed enormous interest in artificial intelligence (AI) due to its military potential. This policy paper examines Chinese perceptions of the military use of AI by studying Chinese-language scholarly materials.
A key finding is that, while the Chinese strategic community has explored the military use of AI across various countries, the United States plays a central and unique role in shaping Chinese perceptions of militarized AI.
The US serves as a global near-competitor, providing a benchmark for China to measure its own development and competitiveness. In addition, the US acts as a role model for the Chinese military to emulate in terms of its ideas, policies and practices.
Consequently, American success in AI has become a primary source of anxiety among the Chinese strategic community, prompting self-reflection and accelerating the development of ambitious Chinese AI plans.
Contrary to the popular narrative in Washington that China has already surpassed the US in the global AI race, Chinese discussions reveal considerable admiration for American AI leadership, with a focus on catching up rather than overtaking the US.
This paper suggests that both the US and China need to play their parts in mitigating the risks of a global AI race.
Artificial intelligence … is recognized as the disruptive technology that is most likely to change the future world … the emergence of AI weapons will fundamentally change the way of warfare. … The world's military powers represented by the United States … have laid out a series of research plans in advance, and released the ‘Third Offset strategy’ in order to widen the gap with potential opponents … China urgently needs to catch up and vigorously develop military application research of AI technology.1
The rise of artificial intelligence (AI) has considerable potential to transform every aspect of our society, including future forms of warfare. The first and second military revolutions were led by the rise of gunpower and nuclear weapons, respectively, and many consider AI as possessing the capacity to launch the third revolution.2 Major military forces—especially those of the US, Russia and China—have been competing to leverage the development of AI for military advantages. Not surprisingly, the growing geopolitical tension between China and the US has accelerated their AI arms race.
Nowadays, one of the most popular concerns in American political narratives and media headlines—if not the most popular—is that China has won the AI race against the US.3 China's unique political system is argued to possess a ‘distinct’ advantage in this race,4 as its AI strategy is often oversimplified by the relevant analyses as a top-down geopolitical masterplan under a national, concerted Chinese effort.5 Many thus call for the US to adopt a similar ‘unified/integrated’ ‘whole-of-nation/government/society’ approach to boost American innovation in AI.6 When it comes to military innovation more specifically, China is considered to be most ‘purposefully’ integrating AI to advance its military,7 posing considerable challenges to American military supremacy. Indeed, those narratives motivated the administration of Joe Biden to sanction and embargo some of the most advanced chips in the hope of slowing down China's AI innovation.8 In other words, these US narratives have already accelerated the US–China AI race by triggering real-world actions.
This policy paper shifts the focus to investigate Chinese perceptions of the US–China AI race. It empirically examines how China perceives AI in the military domain by exploring the following questions. First, in the context of the rapid development of AI across the world, how does China perceive the military use of AI? Second, what role(s) does the US play in shaping this Chinese perception? How does China assess US–China AI competition and reflect on its own AI progress? To answer these questions, the study uses content analysis to study 140 academic Chinese articles on the topic of AI in the military domain.
The study finds that the Chinese strategic community highly regards AI's value in promoting military transformation. According to some Chinese scholars, despite safety and reliability concerns, advancing the military use of AI will be a ‘necessary’ option for armies worldwide and thus AI will reshape the future global military landscape.9 At the same time, the Chinese strategic community acknowledges AI's rapid development in Chinese society but notes slow progress in military innovation. It is argued that the Chinese pace of developing AI in the military domain is ‘relatively lagging behind’, and the development of AI in Chinese military training is ‘relatively weak’.10 The Chinese literature has closely examined the military use of AI in major countries including the US, Russia, Israel, Japan, South Korea and the United Kingdom, studying its implications for China.
Most importantly, this study finds that the US plays a central and unique role in China's strategic discussion. This policy paper analyses the extent to which—and how—the US factor has entered Chinese strategic discussions on the military use of AI. American AI innovation and strategy have influenced the Chinese approach to developing AI in the military domain by promoting Chinese self-reflection and emulation. The US acts as a global near-competitor, if not a geopolitical rival, for the Chinese strategic community to benchmark its own progress and prospects. This self-reflection has often raised strategic anxiety about the widening US–China military gap led by the rise of AI, pointing to the urgency of additional state intervention and support. Precisely because of this strategic anxiety and recognition of American success in developing AI in the military domain, the Chinese military has closely observed and emulated US ideas, policies and practices, further advancing US influence in China's strategic discussion and stimulating Chinese countermeasures. In this regard, from the Chinese perspective, American AI innovation and plans are a source of strategic instability that has widened the bilateral military capability gap and that thus requires Chinese intervention to restore the strategic balance.
Given AI's potential, the risks brought by the global AI race are very high. The ‘race to the bottom’ game in AI, for example, may lead to rapid and uncontrolled AI development, inviting considerable danger to human society.11 An effective global governance system to promote robustness, reliability and safety in military AI has thus become increasingly important nowadays. The nascent global AI governance is yet to prove its worth in mitigating the risks of the US–China AI race. The findings of the study, however, suggest that a window of opportunity lies in taking advantage of the US influence on Chinese perceptions and thus in shaping the latter to buy into safe norms and practices in the military domain of AI.
Methodology and data
In order to study Chinese perceptions, this study employs content analysis of Chinese-language materials. Data is sourced from the China National Knowledge Infrastructure (CNKI), China's largest academic database. All articles in CNKI containing both ‘intelligent’ (zhineng) and ‘military’ (junshi) in the title were selected during the data collection, which process resulted in a primary focus for this study of 140 Chinese articles, published between 1988 and 2022. This period of interest captures China's early exploration of military AI and the first wave of major AI strategy development.
Each of the 140 articles—in total, over 800,000 Chinese characters—was reviewed. Weiciyun, Chinese-language text analysis software, was used to quantitatively count word frequencies and visualize the results (see figure 2) and to analyse word sequences in the text (see figure 3).12 Discussions on the military use of AI span interdisciplinary conversations involving a wide range of academic subjects. The authors of these 140 articles mainly include Chinese military officers, defence industry engineers, policy analysts and academics, most of whom are affiliated with Chinese military institutes, universities and companies. Generally, the scholarly articles studied reflect a Chinese strategic discussion on the military use of AI, involving participants from both academic and military backgrounds. Since all articles are openly accessible resources, information pertaining to secretive military intelligence or internal documents was not examined in this study. Consequently, the discussions analysed do not represent the Chinese military's official stance, but provide insights into it.
The US factor in China's strategic discussion
The Chinese strategic discussion highly values the military utility of AI, with a consensus emerging that AI will profoundly influence future warfare.13 Similar to the ‘killer robot’ narrative, some Chinese scholars even suggest that this will fundamentally alter the nature of warfare, shifting from wars among humans to scenarios where ‘machines automatically kill humans’.14 Such perspectives are hardly surprising, given the data sample. The strategic discussion on AI in the military domain typically involves proponents of its potential. However, a few express cautions regarding its military applications. For example, Li Hao and Du Yanbo criticize views that overestimate AI's impact, cautioning against a hasty adoption of AI in the activities of the People's Liberation Army (PLA). According to Li and Du:
[If we] develop military theories and promote reform practices based on a vague understanding of AI prospects and maximized imagination, it will not only be useless but even detrimental to the transformation and development of military modernization.15
Li and Du conclude that the introduction of AI may bring ‘inefficiency, confusion, and disorder’ and may thus prove ‘counterproductive’ to China's military transformation.16 Nonetheless, this caution is rarely reflected in the 140 analysed articles, which overwhelmingly advocate urgent actions for China to take to embrace the age of AI.
When examining the discussion among Chinese writers, this study finds that the US factor has played a pivotal role from the outset. Figure 1 illustrates the number of publications on the topic, revealing that the first Chinese writing on AI development in the military domain took place in 1988. The initial article, by Qian Xiaojiang, focused on the AI military decision support system (AIMDSS), with some discussion on its application in the US.17 In light of the rapid development of the AIMDSS in the latter state, the author suggested that China should ‘have some sense of urgency’ and begin studying AIMDSS itself. However, the Chinese defence community maintained minimal interest in this topic until 2015.
Number of Chinese articles published annually on militarized AI (1988–2022)
The years 2015 and 2016 marked the onset of China's rapidly growing interest on the topic of militarized AI, coinciding with a series of AI-related events and successes in the US.18 The introduction of the Third Offset strategy by then-US Secretary of Defense Chuck Hagel at the Reagan Defense Forum in late 2014 garnered close attention from China's defence community. Many Chinese strategists viewed the strategy as leveraging AI to accelerate combat capacity, thereby enhancing US conventional defence19 or maintaining its military advantage over China and Russia.20 The consensus was that China needed to review its own military strategy amid the transformation spurred by emerging technology like AI.21 As one Chinese article pointed out:
The world's military powers, represented by the US, foresee the broad application prospects of AI technology in the military field. They believe that the future arms race is an intelligent race and have laid out a series of research plans in advance, releasing the Third Offset strategy and striving to widen the gap with potential opponents in terms of intelligence. Being backward means being under the control of others. To avoid the huge generation gap caused by the loss of development opportunities, China urgently needs to catch up and vigorously develop military application research of AI technology.22
In early 2016 the success of Google DeepMind's AlphaGo in defeating top-ranked human players—including Ke Jie of China and Lee Se-dol of South Korea—in the ancient Chinese board game of Go further shocked not only China's defence community but also the general public. AI became a widely discussed public topic across China. Some international analysts considered it to be a ‘Sputnik moment’ for China that spurred funding and stimulated China's AI ambition.23 Despite being developed in a civilian context, China's defence community clearly recognized AlphaGo's military potential. Among the 140 articles in this study, the most frequently cited and downloaded article is ‘Analysis of AlphaGo's technical principles and prospects for the military application of artificial intelligence’, focusing on the technical principles of AlphaGo and its military potential.24 This article, supported by the Shared Army Major Research Plan Joint Fund and the National Natural Science Foundation of China, concluded that AlphaGo's success had the potential to revolutionize AI-enabled military decision-making. The author of the article, Hu Xiaofeng from the PLA's National Defence University, concluded in a separate article that:
The development of Deep Green and the breakthrough of AlphaGo have given us a lot of inspiration. We can't be eager for quick success, but we can't sit still … Some people say that in the past, ‘mechanization has not caught up, informatization is desperately catching up, and intelligentization can no longer lag behind’. The author believes that mechanization is easier to catch up through equipment update. Informatization requires time accumulation and scale effects, but it can also gradually catch up. Intelligentization, however, is not the same. There are essential differences between them. Once we can't catch up, we may never catch up, because war may not give you time to catch up.25
In these respects, both the Third Offset strategy and AlphaGo's success engendered an inherent sense of insecurity within China, compelling it to react.
In 2016, the same year as AlphaGo's success, the administration of Barack Obama released a report titled Preparing for the future of artificial intelligence, together with a companion National artificial intelligence research and development strategic plan.26 These two reports were viewed by many Chinese strategists as signals of the US' intention to elevate its AI plans to a national strategy.27 Consequently, China swiftly emulated this approach. In July 2017 the Chinese State Council introduced China's national AI strategy, the New generation artificial intelligence development plan. Some international analysts noted considerable similarities between this Chinese strategy document and its US counterpart, from top-level strategic objectives to policy details and recommendations. It was argued that ‘China is embracing and implementing America's [AI] strategy’.28
Envisioning China's ambition to become an AI superpower by 2030, the Chinese State Council anticipates the expansive integration of AI into national defence construction.29 The New generation artificial intelligence development plan considers ‘the deep integration of artificial intelligence with the economy, society and national defense’ as a primary objective.30 Specifically, the plan aims to leverage AI to support ‘military command and decision-making, military deduction, defense equipment, and other applications’. The military–civilian integration concerning AI is also identified as one of the plan's six key tasks.
The growing volume of Chinese writings on the military use of AI thus coincided with the publication of China's national AI strategy and concerns about the widening gap in military capability enabled by AI. From 2016 many articles focused on AI's military use in other countries, particularly the US. Among the articles studied, 64.2 per cent—90 articles—mentioned the US. Analysis of the full text of the 140 Chinese articles using Weiciyun revealed that ‘US’ was one of the most frequently mentioned words in the Chinese strategic discussion. Figure 2 shows the top 100 most frequently mentioned words among all the articles; the word ‘USA’ (highlighted in black in figure 2) is among the top 20 most frequently mentioned words, suggesting that the US is a frequent reference point in this Chinese strategic discussion.
Top 100 most frequently mentioned words among 140 Chinese articles studied
More specifically, American practices, ideas, and even terminology were closely observed and discussed. Figure 3 shows the most frequent words to appear immediately before and after the word ‘US/USA/America’ is mentioned. The most frequently mentioned phrases include ‘American AI’, ‘American research’, ‘American national defence’, ‘American navy’, ‘US military’, ‘the US publishes’, ‘the US proposes’, ‘American nation’, ‘American combat’ and ‘the US and Russia’. A typical line, for example, might be: ‘in 2016, the US military proposed the third Offset Strategy with AI as the key supporting technology, released the National AI research and development strategic plan and FY2009–2034 Unmanned Systems Integrated Roadmap, and proposed a series of new operational concepts such as “distributed operations” and “bee swarms” to promote the deployment of related AI military application projects such as “intelligent missiles” and “unmanned autonomous aerial refueling”’.31 All of these point to the US as a critical factor not only in producing Chinese strategic anxiety but also in serving as a role model for China to emulate.
Most frequently mentioned words before and after the word ‘USA’ among 140 Chinese articles studied
To provide a context for comparison, the frequency of mentions of different countries in the Chinese strategic discussion is illustrated in figure 4. It indicates that members of the Chinese strategic community have examined AI military practices in a wide range of countries. The most frequently mentioned countries/regions are the US, Russia, Europe, Japan, the UK, Israel, India, France, Germany, and South Korea.
Frequency of mentions of different countries across 140 Chinese articles studied
Some Chinese scholars even argue that China is lagging behind not only the US, but also Israel and Russia, when it comes to the practical application of AI for military purposes.32 Subsequently, the focus of the discussion is on what China can learn from practices in those countries. This finding aligns with the wider picture of Chinese military evaluations, indicating that future warfare has been evolving from ‘informatized’ to ‘intelligentized’.33 As such, Chinese strategists have widely acknowledged AI's importance and have carefully studied its development in other countries. As a Chinese article points out:
The speed and effect of advancing the military application of AI will profoundly affect the outcome of future wars, and it will also affect a country's international status. In this context, our country should continue to track the latest developments in the field of AI by foreign militaries and learn from the experience and lessons of other countries in promoting the military application of AI.34
Nonetheless, in comparison, the US received the most attention, with its word frequency exceeding that of all other countries combined, suggesting a US-centric discussion.
Conclusion
This policy paper has examined how the United States' AI strategy and progress stirred anxiety and self-reflection among the Chinese strategic community in the period to 2022. From the Third Offset strategy to the White House's reports on Preparing for the future of artificial intelligence and the accompanying National artificial intelligence research and development strategic plan, the American AI strategy has been carefully observed and emulated by the Chinese strategic community, contributing to the development of China's own AI strategy.
As China's AI near-peer competitor and geopolitical rival, American military AI practices serve as a benchmark for evaluating China's progress among the Chinese strategic community. With growing US–China tensions, the rapid development of AI-enabled military practices has fuelled considerable anxiety in China. From the technological breakthrough of Google DeepMind's AlphaGo to the birth of ChatGPT,35 AI progress in the US in both the civil and military spheres has repeatedly shocked China's strategic community, leading to reflections on its own perceived ‘backwardness’. Therefore, strategic anxiety in the context of the AI race is no less significant in Beijing than in Washington. While both sides consider the other a source of strategic instability and call for increased intervention, the difference between them lies in the fact that the aforementioned US narratives are primarily concerned with losing the AI race in general, while Chinese literature focuses—at least for now—on narrowing the technological gap between the US and China in the military arena. Strategic anxieties and a sense of insecurity stemming from these AI race narratives have motivated both countries to intensify investments in building their respective AI military capacity. In this regard, this security dilemma may escalate the global AI race, potentially leading to AI-enabled security conflicts between two nuclear states becoming a self-fulfilling prophecy.
As a group of top AI experts and scientists rightly point out, similar to nuclear weapons, AI represents an existential threat to humanity.36 When AI intersects with nuclear weapons, this ‘extinction’ risk is a step closer, given the impact of AI on strategic stability and nuclear risks.37 Indeed, AI has been a part of the nuclear weapon infrastructure for decades. The future development of machine learning in providing better nuclear intelligence, enhancing nuclear command and control architecture and even making independent nuclear strike decisions has the potential to alter nuclear strategy and deterrence, thus affecting the strategic balance of nuclear power. While delegating independent nuclear command and control authority to AI seems unlikely at the present time, AI has already been integrated into many aspects of nuclear weapon systems. Responsible management of nuclear weapon systems in the age of AI is crucial to human survival.
In this context, transnational cooperation is key to mitigating the risks of military AI. Nuclear states need to develop a set of international regulations to manage the use of AI in nuclear infrastructure and command, and regulation of lethal autonomous weapons systems (LAWS) needs effective global treaties to address legal and ethical concerns. Such developments require transnational cooperation to form a global consensus and take global actions. Unfortunately, in the shadow of the global AI race, narratives focused on competition and rivalry have not only hindered the urgency of transnational cooperation but have also infiltrated nascent global AI governance.
As Roberts et al. point out, global AI governance faces both first- and second-order cooperation problems. The first stems from interstate geopolitical competition within an anarchic global order, while the second arises from institutional dysfunction, complicating efforts to establish effective global AI governance mechanisms.38 Arguably, the US–China AI race has not only problematized first- but also second-order cooperation challenges. The increasingly intensified US-Chinese geopolitical competition over AI has not been restricted to direct interstate competition, with actions such as the American semiconductor ban on China or the Chinese state's efforts to securitize AI.39
The US–China race has indeed brought in considerable second-order cooperation challenges by developing and supporting new global AI governance mechanisms for geopolitical gain. For example, the Global Partnership on AI, an international initiative to advocate responsible AI use based on democratic principles, was also viewed by many in the US as a helpful geopolitical tool against China.40 The impact of geopolitical rivalry is most obvious when it comes to security- and military-focused AI governance initiatives. The AI Partnership for Defense was initiated by the US Department of Defense in 2020 to shape responsible military use of AI using democratic value-based governance in defence matters.41 As then-US Secretary of Defense Mark Esper stated, it aimed to work with US allies to counter China's AI-enabled ‘21st-century surveillance state’ and Russia's ‘AI-and-autonomy-enabled command-and-control schemes’.42 As such, these initiatives are designed as geopolitical tools to defeat China, rather than mechanisms for transnational cooperation. All these factors have contributed to the ‘weak regime complex’ in the current global AI governance landscape, which features a ‘polycentric’ structure with some institutional linkages, but where work remains largely siloed.43
Without effective global governance, the AI arms race poses considerable risks not only to the loser but also to the winner. The pressure of winning the race relies on breakneck speed in building up AI capacity in the military arena, which is likely to be made at the expense of essential research on AI safety.44 For example, in order to gain a first-mover advantage, states may choose to shorten the testing phase of AI-enabled weapons and deploy them before fully understanding their impact.45 In other words, the AI race may end up driving the deployment of powerful but unsafe AI-enabled military technology, posing risks to all stakeholders. If this risk-taking approach is perceived as successful, the competitive security pressure will motivate more states to join this kind of cross-national emulation. This will not only enhance the overall risk for all but also reduce the first-mover advantage and thus end up in no one side's favour.
In this context, both the US and China need to downplay zero-sum thinking and realize their overlapped interest in promoting robustness, reliability and safety in military AI systems and reducing strategic risks. As the findings of this study show, China is interested in understanding not only US technical progress, but also American ideas and policies to advance its AI. After all, the US–China AI race is not only an arms race but also a leadership contest. The competition lies not only in technical advancements, but also in ideas and policies to govern the use of AI. From its repeated calls for regulating the military and general use of AI to its desire to lead and shape future AI norms and standards,46 the Chinese government has signalled its diplomatic and rhetorical commitment47—regardless of its inconsistency and consideration of its own national interest—to the international community. This is clearly a progressive move, and the question lies in to what extent China's actions can match its words.
At the same time, the US needs to be more confident about its AI leadership and act as a leader for collective goods instead of being misled by unnecessary strategic anxiety. Whatever the US does in terms of its AI strategy, policies and regulations has been closely observed and often emulated by other states, especially China. This is where AI leadership in the US can and should exert its influence. American norms and principles—on issues from keeping a ‘human-in-the-loop’ approach with regard to nuclear infrastructure and command, to supporting global AI governance initiatives and to regulating LAWS—have affected China's position for better or worse. How the US builds a robust, reliable and effective military AI sets an example for China and many others. As such, Chinese admiration of American AI leadership can be turned into a force for good, if American pioneering norms and practices are worthy of being followed and admired in the long term.
Footnotes
Wang Li from China Electronics Technology Group Corporation Limited, in an excerpt from one of 140 articles by Chinese scholars examined in this study. Li Wang, ‘Rengong zhineng zai junshi lingyu de shentou yu yingyong sikao’ [Penetration and application of artificial intelligence in the military field], Keji Daobao [Science & Technology Review] 35: 15, 2017, pp. 15–19.
Kai-Fu Lee, ‘The third revolution in warfare’, The Atlantic, 11 Sept. 2021, https://www.theatlantic.com/technology/archive/2021/09/i-weapons-are-third-revolution-warfare/620013. This view is shared by some of the Chinese articles included in this study. For example, see Wenhao Li and Kun Long, ‘Rengong zhineng gei junshianquan dailai de jiyu yu tiaozhan’ [Opportunities and challenges brought by artificial intelligence to military security], Xinxi anquan yu tongxin baomi [Information Security and Communications Privacy] vol. 5, 2021, pp. 18–26. However, there are cautious views suggesting that this will not happen in the near future. See Tom Stefanick, ‘Why the AI revolution hasn't swept the military’, Brookings, 6 May 2020, https://www.brookings.edu/articles/why-the-ai-revolution-hasnt-swept-the-military. (Unless otherwise noted at point of citation, all URLs cited in this article were accessible on 9 Jan. 2025.)
For those who believe that China has already won, please see Heather Long, ‘In Davos, U.S. executives warn that China is winning the AI race’, Washington Post, 23 Jan. 2019, https://www.washingtonpost.com/business/2019/01/23/davos-us-executives-warn-that-china-is-winning-ai-race; Guy Faulconbridge ‘China has won AI battle with U.S., Pentagon's ex-software chief says’, Reuters, 11 Oct. 2021, https://www.reuters.com/technology/united-states-has-lost-ai-battle-china-pentagons-ex-software-chief-says-2021-10-11; Frederick Kempe, ‘The US is falling behind China in crucial race for AI dominance’, CNBC, updated 28 Jan. 2019, https://www.cnbc.com/2019/01/25/chinas-upper-hand-in-ai-race-could-be-a-devastating-blow-to-the-west.html. For those who believe China might win, please see Graham Allison, ‘Is China beating America to AI supremacy?’, The National Interest, 22 Dec. 2019, https://nationalinterest.org/feature/china-beating-america-ai-supremacy-106861; Gregory C. Allen, Understanding China's AI Strategy: clues to Chinese strategic thinking on artificial intelligence and national security, Center for a New American Security, 6 February 2019, https://www.cnas.org/publications/reports/understanding-chinas-ai-strategy; Kempe, ‘The US is falling behind China; Greg Williams, ‘Why China will win the global race for complete AI dominance’, Wired, 16 April 2018, https://www.wired.com/story/why-china-will-win-the-global-battle-for-ai-dominance; Long, ‘In Davos’; Faulconbridge, ‘China has won AI battle with U.S.’; Graham Allison and Eric Schmidt, ‘China will soon lead the U.S. in tech’, Wall Street Journal, 7 Dec. 2021, https://www.wsj.com/articles/china-will-soon-lead-the-us-in-tech-global-leader-semiconductors-5g-wireless-green-energy-11638915759. Others, however, have suggested a different viewpoint: see James Cooper and Kashyap Kompella, ‘No, China is not winning the AI race’, The Hill, 3 Feb. 2022, https://thehill.com/opinion/technology/592270-no-china-is-not-winning-the-ai-race.
Kelley M. Sayler, Artificial intelligence and national security, Congressional Research Service Report R45178 (Washington DC: Congressional Research Service, 2019), https://crsreports.congress.gov/product/pdf/R/R45178/7, p. 23.
Jinghan Zeng, ‘China's artificial intelligence innovation: a top-down national command approach?’, Global Policy 12: 3, 2021, https://www.globalpolicyjournal.com/journal-issue/vol-12-issue-3-may-2021.
Jeffrey Ding, Deciphering China's AI dream (Oxford: Future of Humanity Institute, University of Oxford, 2018), https://www.fhi.ox.ac.uk/deciphering-chinas-ai-dream; Elsa Kania, ‘China's artificial intelligence revolution’, The Diplomat, 27 July 2017, https://thediplomat.com/2017/07/chinas-artificial-intelligence-revolution; Sayler, Artificial intelligence and national security.
Amy J. Nelson and Gerald L. Epstein, The PLA's Strategic Support Force and AI innovation, Brookings, 23 Dec. 2022, https://www.brookings.edu/articles/the-plas-strategic-support-force-and-ai-innovation-china-military-tech.
Alexandra Alper, Karen Freifeld and Stephen Nellis, ‘Biden cuts China off from more Nvidia chips, expands curbs to other countries’, Reuters, 17 Oct. 2023, https://www.reuters.com/technology/biden-cut-china-off-more-nvidia-chips-expand-curbs-more-countries-2023-10-17.
Jun Liu and Wenhua Wang, ‘Rengong zhineng junshi yingyong fazhan shaomiao’ [The development of AI's military application], Junshi Wenzhai [Military Digest] no. 9, 2020, pp. 11–15.
Cunhua Zhuang, Shaojie Liu and Xiepan Wang, ‘Rengong zhineng funeng junshi xunlian’ [AI-enabled military training], Guofang keji [National Defence Technology] 42: 4, 2021, pp. 129–32.
Dan Milmo and Edward Helmore, ‘Humanity at risk from AI “race to the bottom”, says tech expert’, Guardian, 26 Oct. 2023, https://www.theguardian.com/technology/2023/oct/26/ai-artificial-intelligence-investment-boom.
The software is available at https://www.weiciyun.com.
Xiang Li and Tianyu Luo, ‘Rengong zhineng junshi yingyong yuqi guojifa wenti’ [Military application of artificial intelligence and its international law issues], Xinxi anquan yu tongxin baomi [Information Security and Communications Privacy], no. 1, 2021, pp. 99–108.
Wang, ‘Penetration and application of artificial intelligence’; Zhimin Zhang et al., ‘Rengong zhineng zai junshi duikang zhong de yingyong jinzhan’ [Application progress of artificial intelligence in military confrontation], Gongcheng kexue xuebao [Chinese Journal of Engineering] 42: 9, 2020, pp. 1106–18.
Hao Li and Yanbo Du, ‘Liqing rengong zhineng zai zhinenghua junshibiange zhong de bianjie’ [Clarify the boundaries of artificial intelligence in the transformation of intelligent military], Junshi wenzhai [Military Digest], vol. 23, 2022, pp. 7–11.
Li and Du, ‘Clarify the boundaries of artificial intelligence’.
Xiaojiang Qian, ‘Guanyu rengongzhineng junshijuece zhichi xitong de ruogan jishu wenti tantao’ [Discussion on some technical issues of Artificial Intelligence Military Decision Support System], Junshi tongxin jishu [Journal of Military Communications Technology] 27: 3, 1988, pp. 37–42.
Nonetheless, China's various AI initiatives can be traced back to before this period. See Huw Roberts et al., ‘The Chinese approach to artificial intelligence: an analysis of policy, ethics, and regulation. AI & Society, no. 36, pp. 59–77, https://doi.org/10.1007/s00146-020-00992-2; Zeng, ‘China's artificial intelligence innovation’.
Wang, ‘Penetration and application of artificial intelligence’; Zili Ji and Wenhua Zhang, ‘Shijie junshi qiangguo de rengong zhineng junshi yingyong’ [Military applications of artificial intelligence in the world's military powers], Junshi wenzhai [Military Digest], 2020, pp. 7–10.
Huayue Cai and Zhiyuan Wei, ‘Rengong zhineng zai ge junshiqiangguo de fazhan’ [Development of artificial intelligence in major military forces]', Guofang keji [National Defence Science & Technology], 38: 5, 2017, pp. 7–11.
Peter Wood, ‘Chinese perceptions of the “Third Offset strategy”’, China Brief 16: 15, 2016, pp. 1–3, https://jamestown.org/program/chinese-perceptions-third-offset-strategy.
Wang, ‘Penetration and application of artificial intelligence’.
Paul Mozur, ‘Beijing wants A.I. to be made in China by 2030’, New York Times, 20 July 2017, https://www.nytimes.com/2017/07/20/business/china-artificial-intelligence.html.
Jiuyang Tao, Lin Wu and Xiaofeng Hu, ‘AlphaGo jishu yuanli fenxi ji rengongzhineng junshi yingyong zhanwang’ [Principle analysis on AlphaGo and perspective in military application of artificial intelligence], Zhihui yu kongzhi xuebao [Journal of Command and Control] 2: 2, 2016, pp. 114–20.
Xiaofeng Hu, ‘Junshi zhihui xitong zhong de jiqi zhineng: xianzhuang yu qushi’ [Machine intelligence in military command information system: current situation and trend], Xueshu qianyan [Frontiers] no. 8, 2016, pp. 22–34.
Obama White House, ‘The administration's report on the future of artificial intelligence’, blog post, 12 Oct. 2016, https://obamawhitehouse.archives.gov/blog/2016/10/12/administrations-report-future-artificial-intelligence.
Cai and Wei, ‘Development of artificial intelligence in major military forces’; Wang, ‘Penetration and application of artificial intelligence’; Yamei Cai, ‘Rengong zhineng zai junshi lingyu zhong de yingyong jiqi fazhan’ [Application of artificial intelligence in the military field and its development], Zhineng wulian jishu [Technology of Internet of Things and AI] 1: 3, 2018, pp. 41–8.
Gregory Allen and Elsa Kania, ‘China is using America's own plan to dominate the future of artificial intelligence’, Foreign Policy, 8 Sept. 2017, https://foreignpolicy.com/2017/09/08/china-is-using-americas-own-plan-to-dominate-the-future-of-artificial-intelligence.
State Council of China, Guowuyuan guanyu yinfa xinyidai rengongzhineng fazhanguihua de tongzhi [New generation artificial intelligence development plan], 2017, http://www.gov.cn/zhengce/content/2017-07/20/content_5211996.htm.
State Council of China, New generation artificial intelligence development plan.
Ji and Zhang, ‘Military applications of artificial intelligence in the world's military powers’.
Cai and Wei, ‘Development of artificial intelligence in major military forces’, p. 10.
Elsa B. Kania, ‘Artificial intelligence in China's revolution in military affairs’, Journal of Strategic Studies 44: 4, 2021, https://doi.org/10.1080/01402390.2021.1894136.
Xun Zhao, ‘Mei e rengongzhineng junshi yingyong duibi yanjiu’ [A comparative study of military applications of artificial intelligence in the United States and Russia], Guofang keji gongye [Defence Science and Technology Industry], no. 1, 2020, pp. 55–58 at p. 58.
Although the year 2023, when ChatGPT captured China's attention, falls outside the scope of this study.
Thomas Barrabi, ‘Top AI experts warn of tech's “risk of extinction”—similar to nuclear weapons, pandemics’, New York Post, 30 May 2023, https://nypost.com/2023/05/30/ai-experts-warn-of-risk-of-extinction-similar-to-nuclear-weapons.
Vincent Boulanin, ed., The impact of artificial intelligence on strategic stability and nuclear risk: volume I, Euro-Atlantic perspectives (Stockholm: Stockholm International Peace Research Institute, 2019), https://www.sipri.org/publications/2019/research-reports/impact-artificial-intelligence-strategic-stability-and-nuclear-risk-volume-i-euro-atlantic.
Roberts et al., ‘Global AI governance’.
Jinghan Zeng, ‘Securitization of artificial intelligence in China’, The Chinese Journal of International Politics 14: 3, 2021, pp. 417–45, https://doi.org/10.1093/cjip/poab005.
Mia Hunt, ‘US abandons boycott of global AI partnership’, Global Government Forum, 31 May 2020, https://www.globalgovernmentforum.com/us-abandons-boycott-of-global-ai-partnership.
Patrick Tucker, ‘New Pentagon initiative aims to help allies, contractors work together on AI’, Defense One, 9 Sept. 2020, https://www.defenseone.com/technology/2020/09/new-pentagon-initiative-aims-help-allies-contractors-work-together-ai/168343; United States Department of Defense, AI Partnership for Defense (AI PfD) joint statement, 2020.
Tucker, ‘New Pentagon initiative aims to help’.
Roberts et al., ‘Global AI governance’.
Paul Scharre, ‘Killer apps: the real dangers of an AI arms race’, Foreign Affairs, 16 Apr. 2019, https://www.foreignaffairs.com/articles/2019-04-16/killer-apps.
Scharre, ‘Killer apps’.
Jing Cheng and Jinghan Zeng, ‘Shaping AI's future? China in global AI governance’, Journal of Contemporary China 32: 143, 2023, pp. 794–810, https://doi.org/10.1080/10670564.2022.2107391.
Ministry of Foreign Affairs (MFA) of the People's Republic of China (PRC), ‘Quanqiu rengong zhineng changyi’ [Global AI governance initiative], 20 Oct. 2023, https://www.mfa.gov.cn/web/ziliao_674904/1179_674909/202310/t20231020_11164831.shtml; MFA of the PRC, ‘Position paper of the People's Republic of China on strengthening ethical governance of artificial intelligence (AI), updated 17 Nov. 2022, https://www.fmprc.gov.cn/eng/zy/wjzc/202405/t20240531_11367525.html.
Author notes
This article is part of International Affairs' policy papers series—a forum for bringing new insights into policy debates, for rapidly publishing new empirical results and for developing potential solutions to international problems. The author extends sincere thanks to Tingjun Wang for her invaluable research assistance in this study. Deep gratitude is also owed to the reviewers for their insightful comments and constructive feedback during the three rounds of revisions, which have greatly enhanced this work.
