Sage Journals: Discover world-class research

Abstract

The rhetoric of an ‘AI arms race’ between the United States and China has become increasingly prominent over the past 5 years, despite warnings that it is unnecessarily polarising and undermines safe and ethical artificial intelligence (AI) development. However, existing critiques of the AI arms race narrative engage only sparingly with the racialised dimensions of this discourse. In this article, I draw on the rich theoretical insights of Asian American and Asian diaspora studies to show how the AI arms race narrative is deeply racialised in two key ways. First, I show how the rhetoric of an AI arms race builds upon the myth of a ‘Clash of Civilizations’ between the West and the East. This civilisational rhetoric constitutes China and the United States as distinct and mutually opposed cultural entities, thus foreclosing the possibility of more peaceful and cooperative alternatives to the AI arms race. Second, I demonstrate how the US–China AI arms race specifically draws on previous racialised configurations of anti-Asian sentiment, such as techno-Orientalism and the Yellow Peril. I coin the term Yellow Techno-Peril to connote how older European and Americans fears of being overrun or controlled by China are reproduced in the AI arms race. I close by offering recommendations to key stakeholders such as policymakers, decisionmakers, journalists and media organisations as to how they can mitigate and avoid the racialised rhetoric of an AI arms race between the United States and China.

Keywords

Artificial intelligence China Orientalism racism technology United States

Introduction

In 2019, Frederick Kempe, president and CEO of the American think tank Atlantic Council, insisted that ‘the US is falling behind China in a crucial race for AI dominance’ (Kempe, 2019). Similarly, Russian President Vladimir Putin has declared that whoever leads in artificial intelligence (AI) will become the ‘ruler of the world’, while China has announced its plans to become the world leader in AI by 2030 (Haner and Garcia, 2019: 333). These statements reflect the wider phenomenon of AI nationalism, or ‘a new kind of geopolitics’ that connects geopolitical superiority to AI development (Hogarth, 2017). The most prominent example of AI nationalism is the ‘AI arms race’, which frames AI development as a zero-sum game between competitive nation-states (Cave and ÓhÉigeartaigh, 2018). In 2021, the United States (US) National Security Commission on Artificial Intelligence (NSCAI) released its Final Report, which contained its comprehensive strategy for ‘winning’ the AI era. It argued that the US ‘must win the AI competition that is intensifying strategic competition with China’ and that ‘China's plans, resources, and progress should concern all Americans’ (NSCAI, 2021: 2, 16). The NSCAI report epitomises the geopolitical tensions arising in response to AI development, which is frequently framed as an AI arms race between the US and China (Cave and ÓhÉigeartaigh, 2018).

The phenomenon of AI nationalism, and its attendant discourse of the AI arms race, have drawn intense scrutiny from AI ethicists and the arms control/non-proliferation communities (Cave and ÓhÉigeartaigh, 2018; Haner and Garcia, 2019; Roff, 2019; Kak and West, 2023). However, while existing critiques of the AI arms race narrative focus the destabilising impact of race rhetoric on international relations and AI safety, respectively; this article highlights how racial structures of power and racial feeling underpin the concept of a US–China AI arms race. Grounding the contemporary AI arms race between the US and China in longer histories of Orientalism and racial fear, I argue that the US–China AI arms race is the newest iteration of the racial configuration of the ‘Yellow Peril’. I coin the term ‘Yellow Techno-Peril’ to capture how new technological tensions are founded upon and exacerbate existing geopolitical and racial ones. This article does not analyse how AI is used, created or represented in China or any other Asian country. It also does not trace in detail how the rhetoric of an AI arms race shapes Chinese politics, although it is certainly important to consider the global percolation of the AI arms race discourse; take, for example, China laying out a development plan to become the world leader in AI by 2030 (Mozur, 2017). However, this article focuses on how the discourse of the AI arms race between the US and China shapes American coverage of technological and geopolitical developments, and how anti-Chinese sentiment seeps through in this coverage.¹

Specifically, I argue that US media coverage of the AI arms race evokes anti-Chinese racism in three ways. First, it frames the AI arms race as a civilisational or even a biological competition between different civilisational groups, rather than as an economic, technological or political conflict. In doing so, it builds on histories of anti-Chinese racism that portrayed the ‘East’ and the ‘West’ as fundamentally irreconcilable due to their civilisational difference. Second, the AI arms race builds on a history of techno-Orientalist thought and discourse that casts the ‘East’ as both hypo- and hyper-technological. While the locus of where these political and civilisational anxieties have been concentrated has shifted over time, what remains consistent is a pervasive fear that the pace of ‘Eastern’ technological development will outstrip that of the US, resulting in a profound depletion of American power on the world stage. Third, the US–China AI arms race, like other forms of Yellow Perilism before it, frames not just the Chinese state but Chinese people – and specifically, Chinese Americans and Chinese people living within the US borders – as an extension of China's political interests and thus as a threat to the United States. It explores the parallels and the continuities between the US anxieties around Japan's automobile industry in the 1980s and the contemporary fear of China's AI development and deployment. This article then examines how these fears resulted in racial hatred and profiling of Asian Americans, from the 1982 murder of Vincent Chin to the 2018–2022 China Initiative, which aimed to catch Chinese spies in US research institutions.

The proliferation of AI ‘arms race’ rhetoric, 2017–2023

From 2017 to 2023, there has been a noticeable escalation in the framing of US–China technological competition as an ‘AI arms race’. While there are multiple international players in the so-called AI arms race, including Russia, the UK, South Korea, and the European Union (EU), the focal point of the AI arms race is the increasingly tense geopolitical competition between the US and China. As Zwetsloot, Toner and Ding note, ‘The idea of an artificial intelligence (AI) arms race between China and the United States is ubiquitous. Before 2016, there were fewer than 300 Google results for “AI arms race” and only a handful of articles that mentioned the phrase. Today, an article on the subject gets added to LexisNexis virtually every week, and Googling the term yields more than 50,000 hits’ (Zwetsloot, Toner and Ding, 2018). The AI Now Institute's arms race tracker maps the emergence of the AI arms race discourse within the US from 2017, tracing how the language of an AI arms race between the US and China emerged from major technology leaders like former Google CEO Eric Schmidt before spreading to key politicians, policymakers and national bodies (Kak and West, 2023). Major landmarks include the formation of the NSCAI in 2018; the publication of ex head of Google China Kai-Fu Lee's book AI Superpowers: China, Silicon Valley, and the New World Order; the release of the NSCAI's 2021 Final Report, which argues that the US must invest substantially in AI innovation to ‘protect its security, promote its prosperity, and safeguard the future of democracy’; the establishment of the Special Competitive Studies Project (SCSP) in 2021, the successor to the NSCAI; the passing of the bipartisan CHIPS act in August 2022, blocking certain chip exports to China and reducing its capacity to make semiconductors; and President Joe Biden's executive order in August 2023 restricting US investment into Chinese tech (Kak and West, 2023).

Nonetheless, the framing of international AI competition as ‘AI arms race’ has been extensively critiqued for its technological and military inaccuracies, as well as its politically inflammatory character (Roff, 2019; Scharre, 2021; Cave and ÓhÉigeartaigh, 2018; Belfield and Ruhl, 2022). The discourse of the AI arms race draws its social and political potency through reference to historical incidences of technological competition and political polarisation, including the Cold War, the space race and the nuclear arms race; media coverage of the AI arms race repeatedly refers to Alpha Go's 2017 victory over Ke Jie, the number one Go player in the world, as China's Sputnik moment (Lee, 2019), while Frederick Kempe (2019) warns that ‘President Xi had declared a sort of space race or Manhattan Project around AI that is already delivering measurable results’. It is unclear whether drawing these parallels between nuclear proliferation and AI development is helpful; after all, unlike nuclear armaments, AI is not in and of itself a weapon, and it is difficult to disaggregate military AI development from the development and deployment of AI in other spheres of life (especially given that the majority of AI development is driven by the private sector) (Roff, 2019). More importantly, it is also unclear whether the historical example of the nuclear arms race sets a desirable precedent for the AI arms race. As Haydn Belfield and Christian Ruhl note, the proliferation of US nuclear weapons was largely driven by misplaced beliefs that first the Nazis, and then the Soviets, had overtaken the US in nuclear and intercontinental ballistic missiles; these beliefs turned out to be untrue and resulted in the unnecessary proliferation of nuclear weapons (Belfield and Ruhl, 2022). Belfield and Ruhl (2022) consequently warn that ‘policy makers should examine new claims of a “race” in critical technologies dispassionately and rationally and beware of suboptimal arming in response to claims of adversary capabilities… History – especially the history of nuclear competition – shows that such fears can be overblown and costly, and policy makers would do well to remember the cognitive and cultural biases that make people see threats where there are none’.

The escalatory language of the AI arms race is especially dangerous given that it may undermine the safe and ethical development and deployment of AI (Scharre, 2021; Cave and ÓhÉigeartaigh, 2018). Indeed, there are already signs that the AI arms race rhetoric is leading to the intensified pace of AI development; for example, the US Defence department identified the ‘Chinese threat’ as the driving force behind the department's rapid development of specific new AI capabilities. Critics such as Amba Kak and Sarah Myers West (2023) have warned that the arms race rhetoric is frequently weaponised by leaders in Big Tech to encourage closer state–private partnerships and reframe AI policy as industrial policy, further undermining regulatory attempts to control Big Tech monopolies and instate parameters on AI development and deployment. Consequently, prominent figures in AI and Robotics have publicly declared their opposition to the formation of an AI arms race and, in particular, the race to develop autonomous weapons. In 2015, an open letter from the Future of Life Institute which stated that ‘starting a military AI arms race is a bad idea’ was signed by leading AI scientists and roboticists such as Stuart Russell, Barbara J. Grosz and Nils J. Nilsson, as well as other leading figures in science and technology development such as Stephen Hawking, Elon Musk, and Steve Wozniak (Future of Life Institute, 2015). These critiques of both the AI arms race's sensationalist framing, and the potential adverse impact of competitive and unregulated military AI development, suggest that both increasing the international regulation of military AI and decreasing the use of ‘arms race’ rhetoric are essential steps towards mitigating the harmful effects of AI nationalism.

However, these important critiques of the AI arms race rhetoric in US–China relations rarely engage with how structures of race and racism shape and underpin this discourse. In this article I argue that the phenomenon of AI nationalism is inseparable from previous histories of racism and colonialism that have played a central role in the formation of national identities, and continue to drive nationalist forces and ideologies today. I now turn to the core theoretical concepts and ideas about race that drive my analysis of the US–China AI arms race and outline my data collection methods.

Methodology

This article draws on and enriches a rapidly growing body of literature that interrogates the relationship between race and AI. Existing scholarship in AI ethics emphasises how AI both entrenches and creates patterns of racialised discrimination (Broussard, 2023; Hampton, 2021). For example, Safiya Noble (2018) highlights how image labelling and search algorithms reproduce ugly racial stereotypes, while Joy Buolamwini and Timnit Gebru's acclaimed 2018 study Gender Shades foregrounded how four leading automatic gender classifiers performed best on light skinned men and had the highest error rate for dark skinned women (Buolamwini and Gebru, 2018). From the racial history of surveillance and its present-day AI-powered iterations (Browne, 2015) through to the replication of extractive colonial relations (Kwet, 2019), AI clearly reproduces the racial hierarchies of the past and the power structures of the present. Ruha Benjamin (2019) uses the term ‘the New Jim Code’ to account for how AI lends old racial practices a new ‘veneer of objectivity’, repurposing them for an emergent AI age. AI ethicists consequently critique the exploitation of racialised labour in the production of AI, both in terms of the material production of hardware and the accumulation and extraction of personal data (Nakamura, 2014; Browne, 2015; Atanasoski and Vora, 2019; Hampton, 2021); the whiteness of AI (Cave and Dihal, 2020; Katz, 2020); the unequal distribution in the environmental effects of AI, which follow historically established patterns of environmental racism (Kak and West, 2023); AI's reproduction and re-legitimisation of racial pseudoscience (Scheuerman et al., 2021). Ultimately, they consider how AI can be decolonised and whether its decolonisation is even possible (Mohamed, Png and Isaac, 2020; Adams, 2021).

In this article, I bring AI ethics into conversation with a rich body of work from Asian American, Asian diaspora and Chinese diaspora studies to better understand how AI and its attendant discourses are entwined with anti-Asian racism. These fields of study examine the experiences, lives, and knowledge of Asian and Chinese diasporic peoples across the globe. In doing so, they reveal the constitutive role of Asian racialisation in the construction of historical and contemporary racial hierarchies (Lye, 2005; Lew-Williams, 2018). While acknowledging the exacerbation of anti-Asian racism during COVID-19 (Siu and Claire, 2020), these fields foreground how anti-Asian racism ‘is not reducible to hate, and is in fact a persistent, unexceptional presence in the long historical, Civilisational terror-making machine that is the United States’ (Rodriguez, 2021). Moreover, anti-Asian racism and Asiatic racialisation has historically and continues to have an intimate relationship with technology (Morley and Robins, 1995). Asian Americanist scholars have long highlighted how Asian Americans are racialised through their close association with robots, computers and other forms of technology, to the extent that Asian Americans are frequently stereotyped as ‘robotic’ (Bui, 2022; Shah, 2019; He, 2022; Sohn, 2008; Huang, 2019; Schuller, 2017). They have also illuminated how anti-Asian racism has shaped imaginaries of technology development and deployment (Wong, 2017; Nishime, 2017; Sohn, 2008; Huang, 2019; Roh, Huang and Niu, 2015; Lavender, 2017; Bow, 2022; Cheng, 2019). A key area of inquiry here is the cyberpunk genre and its techno-Orientalist aesthetics, with scholars and cultural critics tracing cyberpunk's emergence to the ‘Japan Panic’ of the 1980s (Morley and Robins, 1995; Chun, 2006; Heale, 2009) while simultaneously highlighting its contemporary iterations in science-fiction franchises like Cyberpunk 2077 and Ghost in the Shell (2017) (Yang, 2020). Furthermore, Asian Americanists have foregrounded how Asian American labour historically powered and continues to power Western technological development, from the exploitation of Chinese labour in the construction of the US transcontinental railroad (Karuka, 2019) through to the role of Asian data labellers (Bui, 2022) and Asian diaspora computer scientists and tech workers (Amrute, 2016) in the contemporary AI industry. This historic entwinement of Asianness and technology in the Western racial imagination, I argue, shapes and conditions the US–China AI arms race rhetoric today.

The data for this article was collected over the course of 2022–2023. As a research Fellow for the AI Now Institute, I followed international coverage of the AI arms race between the US and China for the AI Now Institute's arms race rhetoric timeline (Kak and West, 2023). To do so, I set up a daily set of Google alerts for the following topics:²

- AI arms race

- China antitrust US

- China US AI

- China US AI competition

- China US AI privacy

- China US artificial intelligence

- China US big tech

- China US tech competition

Next, I skimmed through all the headlines and identified pertinent articles, before entering them into the AI arms race timeline database, focusing in particular on US-produced media. Building on the rhetorical and discursive turn in Science and Technology Studies (Bareis and Katzenbach, 2022), I provide a critical, qualitative analysis of the language used to describe the AI arms race between the US and China. Drawing on the methodological approaches of feminist political theorists like Iris Marion Young (2003) and Michelle Lazar (2007), I do not primarily approach race and gender as attributes of individual persons; nor does my analysis of how race and racism shapes the US–China AI arms race focus on the racist beliefs of powerful individuals. Instead, I approach race and gender as a set of power structures and affective forces that condition and render sensible what constitutes the ‘political’ (Young, 2003). Feminist critical discourse analysis interrogates the ‘complex, subtle, and sometimes not so subtle, ways in which frequently taken-for-granted gendered assumptions and hegemonic power relations are discursively produced, sustained, negotiated, and challenged in different contexts and communities’ (Lazar, 2007: 142), treating gender as an element of interpretation rather than explanation (Young, 2003). Similarly, in my analysis of US coverage of the US–China AI arms race, I explore how racialised assumptions and hegemonic power relations are produced, sustained, negotiated and challenged.

Results and discussion

In the discussion, I explore how race shapes and configures US media coverage of the US–China AI arms race in two ways. First, I argue that the zero-sum logic of the US–China AI arms race rhetoric does not only lock the two competitors into a dangerous battle for technological dominance. It also casts the AI arms race as a civilizational conflict, or a battle between two mutually opposed and irreconcilable civilisations whose cultural, linguistic and essential differences make cooperation unlikely, if not impossible. Second, I argue that US coverage of the US–China AI arms race reproduces techno-Orientalist tropes about Asian people, casting their Chinese competitors as untrustworthy, devious and brutal. I explore how this unflattering framing directly parallels the US characterisation of Japan during the 1980s ‘Japan panic’ and the emergence of the techno-Orientalist paradigm. I then examine how US coverage of the AI arms race builds on a much longer history of anti-Asian racism that has constituted Asian Americans as ‘perpetual outsiders’. This culminates in a comparison between the 1982 murder of Vincent Chin and the US Department of Justice's 2018–2022 China Initiative.

The US–China AI arms race as a civilisational conflict

The first, and perhaps most striking, trend to emerge from US coverage of the AI arms race between the US and China is its framing of the AI arms race as not only a geopolitical or technological conflict, but as a civilizational conflict. By this, I mean that rather than being construed as a conflict between two geopolitical rivals, media coverage of the AI arms race increasingly frames the race as a competition between two essential different and diametrically opposed civilisations. In a highly influential and controversial Foreign Affairs article titled ‘The Clash of Civilizations’, the political theorist Samuel Huntington argued that the post-Cold War order would primarily consist of clashes between ‘civilisations’, which he defines as ‘the highest cultural grouping of people and the broadest level of cultural identity people have short of that which distinguishes humans from other species’. He asserts that ‘the great divisions among humankind and the dominating source of conflict will be cultural. Nation states will remain the most powerful actors in world affairs, but the principal conflicts of global politics will occur between nations and groups of different civilizations’ (Huntington, 1993: 22). Writing in 1993 after the end of the Cold War, Huntington argues that international politics is moving ‘out of its Western phase, and its centerpiece becomes the interaction between the West and non-Western civilizations and among non-Western civilizations’, with non-Western civilisations moving from ‘the objects of history’ to join the West as ‘movers and shapers of history’ (Huntington, 1993: 23). While previous global political orders were based on economic competition or political competition, Huntington argued that future global conflicts would be ontological, rather than political: ‘in conflicts between civilizations, the question is “what are you?” That is a given that cannot be changed’ (Huntington, 1993: 7). Huntington did not believe that conflicts between civilisations were inevitable but did argue that civilisational conflicts were likely to be more violent and intractable, exacerbated by fundamental cultural and social differences.

Huntington's theory of the clash of civilizations has rightfully been critiqued for promoting forms of racial and civilizational essentialism (Said, 2001) and underplaying the crucial role played by global and transnational flows in shaping national and regional cultures (Fan, 2015). Furthermore, his characterisation of the West as the progressive, modern, masculine, active shaper of history compared to non-Western civilisations as the passive objects of history reproduces Orientalist and colonial ideologies that racialise non-Western peoples and civilisations as feminine, barbaric and premodern (Said, 1979). The racialised essentialism of a civilisational model of conflict significantly undermines attempts to build cross-cultural AI ethics strategies and global AI governance infrastructures; as Whittlestone et al. note, ‘while value differences between cultures certainly exist, claims about how those differences manifest often depend on unexamined concepts and entrenched assumptions, and lack empirical evidence’; they also oversimplify the relationship between ‘Eastern’ and ‘Western’ value systems (ÓhÉigeartaigh et al., 2020: 577). Yet, Huntington's concept of civilizational conflict, especially in the aftermath of the 9/11 terror attacks and the US government's subsequent ‘War on Terror’, has continued to shape US understandings of itself and its role in the current global world order (Brands, 2019). This has laid down the philosophical and political foundations for an AI arms race based on fundamental civilizational difference.

The framing of the AI arms race as a clash between civilisations – rather than merely a geopolitical or an economic competition – emerges from key US policymakers, industry leaders and media commentators. Central to this civilizational discourse is the assertion that the West (led by the US) and China are built on fundamentally distinct and mutually opposed philosophies and moral reasonings, and that these different value systems underpin and shape national technology development strategies. The NSCAI Final Report makes this explicit, claiming that the AI arms race is not just an economic or technological competition but a ‘value competition’ (NSCAI, 2021: 2). While there are, of course, meaningful social, historical, economic and political differences between the US and China, the ‘clash of civilizations’ discourse reduces and simplifies these differences under the homogenising mantle of ‘culture’. Akin to Huntington's clash of civilisations thesis, commentators such as Craig Smith (CEO of Eye.on.AI and journalistic correspondent for the National Security Commission on Artificial Intelligence), attributes China's supposed advantage in the AI arms race to its Confucian heritage and values. Smith writes:

In some ways, China's AI advantage might be traced to Confucius, whose fifth-century BC philosophy still informs the Chinese worldview, including its devotion to a strong social hierarchy. By discouraging individualism and encouraging the preservation of a strong central authority, this worldview has allowed the government a virtual carte blanche in the collection and use of data. And while Chinese citizens may grouse about controls on civil liberties, they accept those controls with a docility that surprises Western observers. (Smith, 2019)

Smith's characterisation of Chinese people as docile and passive in the face of Chinese state policies ignores the meaningful contributions made to AI ethics by Chinese researchers, academics and industry leaders, as well as the Chinese public's increasing contestation of private companies’ encroachment on their data and their privacy (Roberts et al., 2021). It also obscures how the ‘strong central authority’ that Smith attributes to Confucianism may be less a product of an entrenched cultural preference for centralised power and more a product of the distinct economic and political structures of the contemporary Chinese Communist Party and the disciplinary power of its authoritarian state.

Crucially, though, it frames the AI arms race as a competition between two civilisations – one Confucian, one Judeo-Christian – whose distinct cultural characteristics will shape the trajectory of the AI arms race and global AI development and deployment as a whole. Meanwhile, leading US military figures such as three-star Air Force General Lt. Gen. Richard G. Moore Jr argue that the US military AI development is inherently more ethical than that of its geopolitical rivals due to the US Judeo-Christian cultural foundations. In July 2023, Moore stated that ‘regardless of what your beliefs are, our society is a Judeo-Christian society, and we have a moral compass. Not everybody does…There are societies that have a very different foundation than ours’ (Avi-Yonah, 2023). Likewise, former Google CEO, former NSCAI Chair, and current SCSP Chair Eric Schmidt similarly argues that in the era of the AI arms race, AI ‘should be done with American Values’, not Chinese ones (‘Tech Giant Eric Schmidt Warns China is Catching Up to US in AI’ 2021). An ‘obvious example’ of these American Values, Schmidt argues, is that ‘we don’t believe in discrimination. So systems that would automatically discriminate are not good. Things which have inherent biases that we dislike in our country are not ok with us’ (‘Tech Giant Eric Schmidt Warns China is Catching Up to US in AI’ 2021). Schmidt's insistence that anti-discrimination is a distinctly American ideal is particularly ironic, given the US colonial genocides directed against indigenous peoples, its violent enslavement of Black people and its direct involvement in numerous neo-colonial wars in the Global South. While the NSCAI's critiques of the Chinese Communist Party's authoritarian political structures and state violence remain legitimate and pertinent, the language of a values competition between the US and China risks entrenching the AI arms race as an inevitable conflict between two civilisations, one of which, the US, is inherently more morally righteous than the other – and thus exempt from public criticism and ethical oversight.

The polarising quality of the AI arms race as a clash of civilisations is further evidenced by the framing of AI competition as a zero-sum game, where ‘AI superpowers’ are rapidly being locked into an existential competition for global and technological dominance. Jascha Bareis and Christian Katzenbach (2022) note that national AI strategies largely portray the advent of AI as a transformational societal rupture, casting AI as ‘an unforeseen revolution that penetrates every pore of society and makes past reassurances shaky and obsolete’ (p. 864). The SCSP reflects the temporal horizon of the national AI strategies explored by Bareis and Katzenbach, stating, ‘we want to ensure that America is positioned and organized to win the techno-economic competition between now and 2030, the critical window for shaping the future’ (‌SCSP, n.d.). The zero-sum nature of this techno-economic competition, and the shrinking temporal window for ‘shaping the future’, frames any other civilisation's technological development as an inherent threat to the geopolitical supremacy of the US. For example, Kempe (2019) argues that the US needs to win the Fourth Industrial Revolution because ‘countries that are most innovative and technologically advanced tend to dominate international relations’. Consequently, he warns that the US must ensure that China's AI-powered challenge to US leadership does not ‘damage US interests for decades to come and accelerate the decline of US global leadership’ (Kempe, 2019). Will Roper (2020), writing for the online technology publication WIRED, argues that ‘though there is no moon in this digital “space race”, its giant leap may be the next industrial revolution… It's not just our military that needs digital urgency: Our nation must wake up fast. The only thing worse than fearing AI itself is fearing not having it’. The zero-sum logic of the AI arms race between the US and China is particularly evident in a 2017 Pentagon report, which argues that ‘if we allow China access to these same technologies concurrently, then not only may we lose our technological superiority, but we may even be facilitating China's technological superiority’ (Mozur and Perlez, 2017). The framing of technological competition between the US and China as inevitable, intractable and a zero-sum game feeds the ideology of a clash between mutually opposed civilisations – between whom there can only be one victor.

The reframing of AI competition as a zero-sum conflict between the Judeo-Christian US and Confucian China shows how AI development and deployment has become increasingly imbricated with national mythologies. The US national AI strategy and policies embody many of the core ideologies that underpin US identity and politics. Take, for example, the ideology of American exceptionalism; the Trump Administration's US Deputy Assistant for technological development, Michael Kratsios, states, ‘Generation after generation, American innovation has benefited our people and the entire world. American oil fueled world industries. American medicine conquered diseases. [ … ] Today, with so many of the mysteries of quantum computing, autonomous systems, and machine learning yet to be discovered, we can take hold of the future and make it our own’ (Bareis and Katzenbach, 2022: 866). Similarly, Eric Schmidt calls for a return to American exceptionalism in the face of the growing challenge posed by China's AI development, arguing, ‘weren’t we the ones that invented this stuff? Weren’t we the ones that were going to go exploit the benefits of all this technology for betterment and American exceptionalism, in our own arrogant view’ (Vincent, 2017). Consider also the US ideological roots in liberal individualism: the conference proceedings of the 2018 White House Summit on AI for American Industry states that falling behind in the AI arms race would pose ‘growing challenges to societies that have been built upon individual freedom’ (Bareis and Katzenbach. 2022: 868). In a comparative analysis of documents produced from the US, China, France and Germany, Bareis and Katzenbach (2022) argue that ‘in the US version, AI embodies the free spirit of American scientific ingenuity, the dedication of hardworking people in the rust belt, the competitive economic strength of a proud nation building on a long tradition of narratives of progress and America's culture of greatness’ (p. 872). Through these ideologies, the AI arms race is incorporated into a much longer nationalist narrative about American superiority and the US central role in the global order.

Crucially, the American exceptionalism and moralised concept of American values built into the US AI arms race strategy builds on a much longer entwinement of national and civilizational identity with technological prowess. Kratsios’ exultation of American innovation reflects how technological advancement was historically used to legitimate Western colonial expansion and substantiate ideologies of Western racial and civilisational supremacy throughout the 18th and 19th centuries. The historian Michael Adas notes that technological accomplishment served as ‘the criteria by which one could distinguish civilized from barbarian and savage cultures’ and that specific technologies functioned as ‘indicators of the level of development a given society had attained’ (Adas, 1990: 196). Joel Dinerstein (2006) emphasises the centrality of technology to US national identity and ideology, while Stephen Cave and Kanta Dihal (2023) examine how the US fetishisation of technology was inspired by the ‘ideological needs of settler colonialism’(p. 153). The rhetoric and discourse of the AI arms race thus may be driven, in part, by the role that technology plays as a ‘white mythology’ that helps maintain two central myths: ‘the myth of progress’ and ‘the myth of white, Western superiority’ (Dinerstein, 2006: 570–572). Rather than being a manifestation of a clash between two inherently different and mutually opposed civilisations, perhaps the AI arms race is better understood as the latest iteration of racial and (neo)colonial competition.

These racialised narratives of technological prowess and progress emerge overtly through the use of biological metaphors to describe the civilizational conflict of the AI arms race. AI competition is frequently portrayed in national AI strategies and media coverage of the AI arms race as an existential struggle between civilisations for not just supremacy, but survival: ‘no matter if packed in a rhetoric of “catching up”, “defending the pole position,” or scenarios of “brute survival”, capitalist competition about market shares and military strivings for geopolitical hegemony fostered through advancement in AI technology are portrayed as of pivotal importance’ (Bareis and Katzenbach, 2022: 868). The framing of the AI arms race as a ‘survival of the fittest’ uncomfortably echoes the social Darwinist ideologies of the 18th, 19th and 20th centuries (Dinerstein, 2006: 571). Commentators on the AI arms race draw explicit parallels between the race to develop more sophisticated military AI systems with ‘biological examples of competitive co-evolution’ (Scharre, 2021: 126). Paul Scharre writes, ‘biologists often use the metaphor of an arms race to explain “an unstable runaway escalation” of adaptation and counter-adaption that can occur in animals…This broader biological definition of an arms race is more in line with the potential for an escalating “arms race in speed” among nations that leads to greater automation in warfare’ (Scharre, 2021: 126). Scharre is not alone in his portrayal of the AI arms race as a form of animalistic co-evolution, with Kempe (2019) writing that ‘it took a Soviet satellite called Sputnik in 1957, the first ever in space, to awake America's animal spirits and spur our competitive streak’. However, the framing of the arms race through the lens of biological competition risks reproducing scientifically racist, eugenicist ideas of intractable biological differences between different ‘races’, suggesting that these ‘races’ are at different evolutionary stages and/or in competition with one another for a set of finite resources. The AI arms race is thus cast not only as a clash between diametrically opposed civilisations, but also as an existential conflict between two nations, of which only one can prevail.

China as the ‘Yellow Techno-Peril’

The intense polarisation of the AI arms race as a clash between civilisations is further amplified by the distinct racial character of the US–China conflict. The West – and the US, in particular – has long used technological prowess to assume superiority over other nations perceived as ‘static, primitive, passive, Communist, terrorist, or fundamentalist (depending on the era)’ (Dinerstein, 2006: 571). Nonetheless, the AI arms race specifically builds on and reinvents existing anti-Asian racial tropes, such as Orientalism and the ‘Yellow Peril’. While the ‘clash of civilizations’ could extend to a variety of different cultural forms and groupings across the globe, the contemporary US–China AI arms race gains its rhetorical and affective power through reference to the West's longstanding understanding of China – and the East, more broadly – as its constitutive racial ‘Other’. In 1978, the Palestinian academic Edward Said critiqued how Western academics construed the ‘East’ or the ‘Orient’ as backwards, passive and feminine in comparison to the modern, active, masculine West (Said, 1979). Crucially, Said argued that the West's Orientalism revealed far less about the nature of the so-called Orient itself, and much more about the West's own crises of identity and the contradictory conditions of modernity. The Orientalism of the West's interpretation of the East extends to the East's technological production. In 1995, following the ‘Japan bashing’ and ‘Japan panic’ that characterised the US political and economic landscape in the 1980s, Morley and Robins coined the term techno-Orientalism to describe the West's distinct anxieties and desires regarding the spectre of a technologically ascendant Japan (Roh, Huang and Niu, 2015; Morley and Robins, 1995). Central to these anxieties was the notion that ‘Japan is the future, and it is a future that seems to be transcending and displacing Western modernity’ (Morley and Robins, 1995: 168). While the locus of these anxieties has shifted since Japan's asset price bubble collapse in 1991, the affective structures of techno-Orientalism have remained in place. Rather than being solely confined to Japan and Western perceptions of Japaneseness, they have been transposed and translated across to various countries and nations in the ‘Far East’, in particular China.

Furthermore, in the case of the US–China AI arms race, the queasy relations of fetishisation and fear that characterise techno-Orientalism become imbricated in a second, much more overtly hostile, racial trope: that of the ‘Yellow Peril’. The term Yellow Peril describes the West's racialised fears of invasion, domination and imperilment by forces from the so-called ‘East’ (Tchen and Yeats, 2014). Key to the Yellow Peril's racialising force is its conflation of a country's government, its ‘culture’, and its economy with its people, including its extensive diaspora. In 1971, Webster's New International Dictionary defined the Yellow Peril as ‘1. A danger to Western Civilization held to arise from expansion of the power and influence of Oriental people; 2: a threat to Western living standards developed through the incursion into Western countries of Oriental labourers willing to work for very low wages and under inferior working conditions’ (Frayling, 2014: 10). While fears of an Asian uprising had been present in Europe since Genghis Khan, it was amplified by the invention of the ‘yellow race’ in 18th and 19th century race science (Keevak, 2011), populist uprisings against Chinese immigrants (Lew-Williams, 2018), and the proliferation of Yellow Perilist characters and tropes in US popular culture. The tenets of Yellow Perilism were observable in the US political establishment and media's response to Japan's technological development in the 1980s, where Japan's ascendency was portrayed as an existential threat to the economic prosperity and geopolitical supremacy of the US state (Heale, 2009; Garvey, 2020). In the case of the AI arms race, the Yellow Perilism originally directed towards Japan's technological advancement, which at the time was both a democracy and a military ally of the US, is further amplified by the geopolitical and military animosity between the US and China (Garvey, 2020). By casting China's AI development as a fundamental concern for all Americans – as the NSCAI 2021 Final Report does, alongside other key US AI strategy documents – the rhetoric around the AI arms race reproduces previous histories of Yellow Perilism that cast the Chinese state as an inherent threat to the US. In the AI arms race, the techno-Orientalism previously directed toward Japan combines with the much longer Sinophobic history of the Yellow Peril to create a new racial configuration: the Yellow Techno-Peril.

Moreover, the AI arms race re-animates forms of anti-Chinese racism that constitute Chinese people as an ontological extension of the Chinese state, whether or not they have any meaningful political connections to China. The Chinese state is established as the external existential threat to the US, while Chinese-American and Chinese people living within the US borders are constituted as dangerous ‘enemies within’. The rhetoric of the Yellow Techno-Peril does not only further the ideology of a clash of civilisations between two hostile techno-superpowers: it also dehumanises Chinese Americans living within US borders, entrenching the deeply held racial belief that Chinese people are ‘permanent sojourners’ or eternal aliens in the eyes of the US state (Lew-Williams, 2018). There are strong parallels between the current wave of Sinophobic and anti-Chinese rhetoric in the AI arms race and the Yellow Peril politics of the 1980s, which had negative ramifications on Japanese Americans and Asian American communities in the US (Heale, 2009; Garvey, 2020). Take, for example, the 1982 murder of Vincent Chin, which is now recognised as a landmark event in the formation of an Asian American political identity in the US. As Neda Atanasoski and Kalindi Vora write, ‘layoffs in the US auto industry, perceived by the public as a response to the dominance of Japanese automobile manufacturing, led to a series of attacks on Asian Americans. This includes the infamous 1982 beating of Vincent Chin in Detroit by two laid-off Chrysler workers, which resulted in Chin's death…killers justified their attack by saying, “It's because of you we’re out of work”, linking the Asian American Chin to Japanese autoworkers’ (Atanasoski and Vora, 2019: 45). Vincent Chin was not even Japanese, let alone an automobile worker; instead, he was a Chinese American draftsman. Chin's murder demonstrates how easily Yellow Perilist fearmongering moves and slips between different Asian peoples and countries, from the technologically ascendant Japan of the 1980s through to the contemporary AI arms race between the US and China today. Moreover, it also demonstrates how the Yellow Techno-Perilist rhetoric of the time – in particular, the fearmongering that, in the words of US Senator John Dingell, ‘little yellow men’ were taking white US automobile workers’ jobs – became attached to the bodies of Asian Americans, with violent and disastrous results.

The scapegoating of Asian American people as internal threats and permanent aliens who embody white fears of the Yellow Peril extends into the age of the AI arms race, most notably through the US government's China Initiative. In 2018, the Trump administration founded the China Initiative to counter Chinese economic espionage and protect the US national security interests (Guo, Aloe and Hao, 2021). The Trump administration's emphasis on Chinese economic espionage reproduced earlier forms of fiscal orientalism (Yang, 2016) and monetary orientalism (Bui, 2019) that cast China as a ‘barbarian economic aggressor’ (Zhang, 2021) and a ‘global cheater’ (Bui, 2019; Garvey, 2020). Even more troublingly, it gave further credence to forms of racial internal policing that have historically been used to profile and punish Asian Americans as perceived spies. As Frank H. Wu (2023) notes, ‘with conflict between the United States and China increasingly a dominant theme of public discourse, Asian Americans, visible as such, have ended up overrepresented in the prosecution of espionage and associated crimes’ (pp. 191–192). In 2018, the FBI director Christopher Wray argued that the ‘China threat’ should not be seen only as a ‘whole-of-government’ threat, but as a ‘whole-of-society’ threat requiring a ‘whole-of-society’ response (Wu, 2023: 192). Through this lens, all Chinese people are seen as an extension of the Chinese government and its aims, regardless of their own personal political beliefs (Wu, 2023: 192). Donald Trump reinforced this view when he insisted that ‘almost every student’ that comes to the US from China ‘is a spy’ (Wu, 2023: 192). Consequently, an investigation by the MIT Technology Review found that the China Initiative evolved from a narrow initiative focusing on economic espionage to an ‘umbrella term for cases with any connection to China’, and that 88% of the defendants charged were of Chinese heritage (Guo, Aloe and Hao, 2021). The racial profiling of Chinese researchers through the China Initiative reflects how ‘three decades after the fall of the Soviet Union, as China emerges as a new superpower and contests US hegemony, the face of foreign espionage in the West has become Chinese’ (Cheng, 2023).

The ‘whole-of-society’ approach to the threat supposedly posed by the Chinese state and, by extension, Chinese people within the US, has even led to calls for the US government to restrict or entirely ban Chinese students. In April 2021, the Republican Senators Tom Cotton, Marsha Blackburn, and Tommy Tuberville introduced the Secure Campus Act, which aimed to ‘prohibit Chinese nationals from receiving visas to the United States for graduate or post-graduate studies in STEM fields and would ban participants in China's foreign talent recruitment programs and Chinese nationals from taking part in federally-funded STEM research’ (‘Cotton, Colleagues Introduce Bill to Restrict Chinese STEM Graduate Student Visas’ 2021). Tom Cotton justified the Bill on the basis that ‘allowing China unfettered access to American research institutions is akin to granting Soviet scientists access to our critical laboratories during the Cold War. We shouldn’t allow the Chinese Communist Party to exploit the openness of American research institutions any longer. The SECURE CAMPUS Act will help stop Chinese nationals from stealing U.S. technology, which the CCP uses against our own troops and businesses’ (Cotton, Colleagues 2021). Cotton, Blackburn and Tuberville are not alone in their desire to ban or restrict the entry of Chinese STEM students: a Pew Research poll found that 55% of Americans want to instate restrictions on Chinese students entering the country (Zwetsloot and Arnold, 2021). The racial logics underlying the Secure Campus Act mimic those that drove the Chinese Exclusion Acts (1882, 1892, 1902), preventing Chinese nationals from entering the US. Central to the China Initiative and the Secure Campus Act is the assumption that Chinese people are ‘permanently loyal to China’ and ‘racially incapable of becoming American’ (Lew-Williams, 2018: 6). Hence, despite arguments that Chinese students enrich the US economy and STEM research cultures, Chinese students and researchers remain marked by the assumption that they are an extension of the Chinese state.

Despite the shutdown of the China Initiative in 2023 and the failure of the Secure Campus Act, the racial profiling of Chinese Americans and people of Chinese heritage as inherent threats to the US technological supremacy continues to haunt Chinese American researchers and technologists. Writing in Nature, Natasha Gilbert argues that the ‘climate of fear and anxiety’ generated by the China Initiative has not dissipated, and that scrutiny of Chinese scientists has only intensified (Gilbert, 2023).

While acknowledging the initiative's failures, Assistant Attorney General Matthew Olsen argued, ‘if you were looking for general deterrence, it has been achieved in spades – we have terrified the entire research community…Deterrence is about fear’ (Barry and Benner, 2022). Olsen's insistence that a nebulous and widespread fear is a good thing obscures how the racial fear of surveillance and violence has long conditioned Asian Americans’ experiences in the US (Lew-Williams, 2018; Wu, 2023), and how this fear is born unequally by scientists of Chinese descent. Consequently, Chinese scientists and researchers have argued that ‘tensions between the United States and China are likely to increase, but Chinese Americans should not be treated as collateral damage’ (Liu et al., 2023: 804), and ‘urge the federal government to shift away from deterrence and fear and instead focus on the philosophies that have made the United States a global leader in science and technology: openness, diversity, and international collaboration’ (Liu et al., 2023: 804). However, without significant efforts to combat and transform the US portrayal of China as the Yellow Techno-Peril, the fear and anxiety experienced by researchers of Chinese descent is unlikely to dissipate.

Conclusion

In this article, I have explored how American approaches to the ‘AI arms race’ between the US and China draw on historical racial stereotypes, portraying China as the Yellow Techno-Peril. Specifically, I have highlighted how US media coverage and political commentary uses racial and civilizational tropes to constitute the US and China as diametrically opposed hostile civilisations. Through this polarising lens, the US AI development becomes a national security imperative, one that requires immense investment and state support in order to ensure that the US does not fall behind China. Falling behind China, US coverage of the AI arms race suggests, would be an existential to the US not just as a state, but as a civilisation. AI development and deployment thus becomes moralised and entangled with American values and ideologies of American exceptionalism and superiority. Furthermore, the US AI arms race rhetoric does not only portray China as an oppositional civilisation that exists in tension with US interests and values. It also draws on older forms of techno-Orientalism and anti-Asian racism to portray China as a distinctly racialised and Othered civilizational threat to the US. This refracts within the US national borders, leading to the racial surveillance and profiling of researchers and scientists of Chinese descent. Constituted as permanent outsiders within the US nation, the AI arms race rhetoric only serves to further alienate Chinese American and Chinese diasporic people in the US. Based on the analysis in this article, I offer the following recommendations to journalists, media outlets, policymakers, politicians and academics who are reporting or researching the AI arms race.

First, commentators and decisionmakers should question whether the language of an ‘arms race’ is appropriate, useful or necessary. As explored in the opening of this paper, it is unclear whether describing AI competition as an ‘arms race’ effectively captures the nature or the geopolitical implications of the AI-powered products being developed and deployed. It can also fuel the belief that an AI-powered future is an inevitability, rather than a conscious social, political and economic choice. We have seen some recent examples of global AI safety cooperation, such as the UK inviting China to its Global AI safety summit and China's Global AI Governance initiative. Nonetheless, it is essential that we continue to find alternative narratives to the AI ‘arms race’ rhetoric to promote better and safer AI development and governance.

Second, commentators and decisionmakers should avoid excessively polarising language, especially rhetoric that positions the US and China as irreconcilable civilisations. It is important and necessary to critique other countries’ AI strategies and policies, particularly in the case of large-scale human rights abuses like the Chinese government's extensive and documented state violence against Uyghur Muslims in Xinjiang. However, rather than attributing these policies to inherent cultural difference, commentators and decisionmakers should focus on how these policies have emerged from political and economic incentives and pervasive racial ideologies. This is also crucial for promoting cross-cultural cooperation and international AI governance.

Third, commentators and decisionmakers must be cognizant of how their language draws on and reproduces previous histories of racial violence and anti-Asian discourses. They must carefully consider how their language or policies may be conflating Chinese people – whether within the borders of the Chinese state, or in the diaspora – with the Chinese government. Framing Chinese people as an inherent extension of the Chinese nation state's politics and national interests risks amplifying anti-Asian sentiment and encouraging acts of anti-Asian violence. This is a particular concern given the rise in anti-Asian hate crimes during the COVID-19 pandemic and the continued deterioration of US–China relations.

Finally, there is an urgent need for further research on the underlying racial structures that shape the geopolitics of AI and the implications of AI development on international relations. The insights of feminist international relations, critical race theory, and postcolonial theory – among other disciplines – are crucial if we are to understand how AI may entrench and deepen inequalities on the global stage. This analysis of the AI arms race is only the first step in this wider intellectual project.

Footnotes

Acknowledgements

Many thanks to Dr Eleanor Drage and Os Keyes for their comments on draft versions of this article. I would also like to extend my thanks to the AI NOW Institute and Dr Stephen Cave for their feedback and support of my work on AI nationalism and the AI arms race.

Author Note

Versions of this paper were presented at the Leverhulme Centre for the Future of Intelligence, the book launch of Xine Yao's Disaffected (a recording of which is available ), and 4S. Part of the paper was also presented at RightsCon, and a related essay on AI Nationalism was published as part of AI Now's New AI Lexicon project. The essay is available here, and has been self-cited throughout the essay.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: the author’s current postdoctoral position is funded by the Stiftung Mercator Foundationand the previous postdoctoral position was funded by Christina Gaw.

ORCID iD

Kerry McInerney

Notes

References

Adams

(2021) Can artificial intelligence be decolonized? Interdisciplinary Science Reviews 46(1-2): 176–197.

Adas

(1990) Machines as the Measure of Men: Science, Technology and Ideologies of Western Dominance. Ithaca: Cornell University Press.

Amrute

(2016) Encoding Race, Encoding Class: Indian IT Workers in Berlin. Durham: Duke University Press.

Atanasoski

Vora

(2019) Surrogate Humanity: Race, Robots, and the Politics of Technological Futures. Durham: Duke University Press.

Avi-Yonah

(2023) Judeo-Christian roots will ensure U.S. military AI is used ethically, general says. The Washington Post, July 22. Available at: https://www.washingtonpost.com/national-security/2023/07/22/air-force-general-ai-judeochristian/ (accessed 21/09/2023).

Bareis

Katzenbach

(2022) Talking AI into being: The narratives and imaginaries of national AI strategies and their performative politics. Science, Technology, & Human Values 47(5): 855–881.

Barry

Benner

(2022) US drops its case against MIT scientist accused of hiding China links, New York Times, January 20. Available at: https://www.nytimes.com/2022/01/20/science/gang-chen-mit-china-initiative.html (accessed 26/09/2023).

Belfield

Ruhl

(2022) Why policy makers should beware claims of new ‘Arms Races’ Bulletin of the Atomic Scientists, July 14. Available at: https://thebulletin.org/2022/07/why-policy-makers-should-beware-claims-of-new-arms-races/ (accessed 21/10/2022).

Benjamin

(2019) Race after Technology: Abolitionist Tools for the New Jim Code. Cambridge: Polity Press.

10.

Bow

(2022) Racist Love: Asian Abstraction and the Pleasures of Fantasy. Durham: Duke University Press.

11.

Brands

(2019) ‘Clash of Civilizations’ has no place in U.S. Foreign Policy, Bloomberg, May 4. Available at: https://www.bloomberg.com/view/articles/2019-05-04/-clash-of-civilizations-has-no-place-in-u-s-foreign-policy?leadSource=uverify%20wall (accessed 21/09/2023).

12.

Broussard

(2023) More than a Glitch: Confronting Race, Gender, and Ability Bias in Tech. Cambridge: MIT University Press.

13.

Browne

(2015) Dark Matters: On the Surveillance of Blackness. Durham: Duke University Press.

14.

Bui

(2019) Monetary orientalism: Currency wars and the framing of China as global cheater. Global Society: Journal of Interdisciplinary International Relations 33(4): 479–498.

15.

Bui

(2022) Model Machines: A History of the Asian as Automaton. Philadelphia: Temple University Press.

16.

Buolamwini

Gebru

(2018) Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. Proceedings of Machine Learning Research 81: 1–15.

17.

Byler

(2021) In the Camps: China's High-Tech Penal Colony. New York: Columbia Global Reports.

18.

Cave

Dihal

(2020) The whiteness of AI. Philosophy and Technology 33: 685–703.

19.

Cave

Dihal

(2023) Fiery the angels fell. In: Cave

Dihal

(eds) Imagining AI. Oxford: Oxford University Press, 149–167.

20.

Cave

ÓhÉigeartaigh

(2018) An AI race for strategic advantage: Rhetoric and risks. AIES: 36–40. https://doi.org/10.1145/3278721.3278780.

21.

CBS News (2021) Tech Giant Eric Schmidt warns China is catching up to U.S. in AI, CBS News, April 21. Available at: https://www.cbsnews.com/news/tech-giant-eric-schmidt-warns-china-is-catching-up-to-u-s-in-a-i/ (accessed 1/3/2022).

22.

Cheng

(2019) Ornamentalism. Oxford: Oxford University Press.

23.

Cheng

(2023) The All-American Myth of the TikTok Spy, WIRED, August 9. Available at: https://www.wired.com/story/china-espionage-tiktok-spying-national-security/ (accessed 27/09/2023).

24.

Chun

WHK

(2006) Control and Freedom: Power and Paranoia in the Age of Fiber Optics. Cambridge: MIT Press.

25.

Cotton Colleagues (2021) Introduce bill to restrict Chinese STEM graduate student visas, Tom Cotton, April 22. Available at: https://www.cotton.senate.gov/news/press-releases/cotton-colleagues-introduce-bill-to-restrict-chinese-stem-graduate-student-visas (accessed 18/12/2023).

26.

Dinerstein

(2006) Technology and its discontents: On the verge of the posthuman. American Quarterly 58(3): 569–595. http://www.jstor.org/stable/40068384 .

27.

Fan

(2015) Techno-Orientalism with Chinese characteristics: Maureen F. McHugh’s China Mountain Zhang. Journal of Transnational American Studies 6(1): 1–33.

28.

Frayling

(2014) The Yellow Peril: Dr Fu Manchu and the Rise of Chinaphobia. London: Thames & Hudson.

29.

Future of Life Institute (2016) Autonomous Weapons Open Letter: AI& Robotics Researchers. Future of Life Institute, February 9. Available at: https://futureoflife.org/open-letter/open-letter-autonomous-weapons-ai-robotics/#:∼:text=In%20summary%2C%20we%20believe%20that,weapons%20beyond%20meaningful%20human%20control (accessed 20 January 2024).

30.

Garvey

(2020) “AI for social good” and the first AI arms race: Lessons from Japan’s fifth generation computer systems project. The 34th Annual Conference of the Japanese Society for Artificial Intelligence: 1–4.

31.

Gilbert

(2023) China Initiative's shadow looms large for US scientists. Nature 615(7951): 198–199.

32.

Guo

Aloe

Hao

(2021) The US crackdown on Chinese economic espionage is a mess. We have the data to show it, MIT Tech Review, December 2. Available at: https://www.technologyreview.com/2021/12/02/1040656/china-initative-us-justice-department/ (accessed 27/09/2023).

33.

Hampton

(2021) Black feminist musings on algorithmic oppression. FAccT ‘21: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency: 1–12.

34.

Haner

Garcia

(2019) The artificial intelligence arms race: Trends and world leaders in autonomous weapons development. Global Policy 10(3): 331–337.

35.

(2022) The racial interface: The computational origins of minority modeling. NMC Media-n 18(1): 31–52.

36.

Heale

(2009) Anatomy of a scare: Yellow Peril politics in America, 1980–1993. Journal of American Studies 43(1): 19–47. http://www.jstor.org/stable/40464347 .

37.

Hogarth

(2017) AI nationalism. Available at: https://www.ianhogarth.com/blog/2018/6/13/ai-nationalism (accessed 30/3/2021).

38.

Huang

(2019) The posthuman subject in/of Asian American literature. In: Rabinowitz, P. (n.d.) Oxford Research Encyclopedia of Literature. Oxford: Oxford University Press, 1–20.

39.

Huntington

(1993) The clash of civilizations? Foreign Affairs 72(3): 22–49.

40.

Kak

Myers-West

(2023) Confronting Tech Power, AI Now. Available at: https://ainowinstitute.org/2023-landscape (accessed 27/09/2023).

41.

Karuka

(2019) Empire's Tracks: Indigenous Nations, Chinese Workers, and the Transcontinental Railroad. Oakland: University of California Press.

42.

Katz

(2020) Artificial Whiteness: Politics and Ideology in Artificial Intelligence. New York: Columbia University Press.

43.

Keevak

(2011) Becoming Yellow: A Short History of Racial Thinking. Princeton: Princeton University Press.

44.

Kempe

(2019) The US is falling behind China in crucial race for AI dominance, CNBC, January 26. Available at: https://www.cnbc.com/2019/01/25/chinas-upper-hand-in-ai-race-could-be-a-devastating-blow-to-the-west.html (accessed 18/2/2022).

45.

Kwet

(2019) Digital colonialism: US empire and the new imperialism in the global south. Race & Class 60(4): 3–26.

46.

Lavender

(2017) Dis-Orienting Planets: Racial Representations of Asia in Science Fiction. Jackson: Mississippi.

47.

Lazar

(2007) Feminist critical discourse analysis: Articulating a feminist discourse praxis. Critical Discourse Studies 4(2): 141–164.

48.

Lee

(2019) AI Superpowers: China, Silicon Valley, and the New World Order. New York: Harper Collins.

49.

Lew-Williams

(2018) The Chinese Must Go: Violence, Exclusion, and the Making of the Alien in America. Cambridge: Harvard University Press.

50.

Liu

Luo

, et al. (2023) US “China initiatives” promote racial bias. Science 380(6647): 804–804.

51.

Lye

(2005) America's Asia : Racial Form and American Literature, 1893-1945. Princeton: Princeton University Press.

52.

Mohamed

Png

M-T

Isaac

(2020) Decolonial AI: Decolonial theory as sociotechnical foresight in artificial intelligence. Philosophy and Technology 33: 659–684.

53.

Morley

Robins

(1995) Spaces of Identity: Global Media, Electronic Landscapes and Cultural Boundaries. London, New York: Routledge.

54.

Mozur

(2017) Beijing wants A.I. to be made in China by 2030, The New York Times, July 20. Available at: https://www.nytimes.com/2017/07/20/business/china-artificial-intelligence.html (accessed 6/12/2023).

55.

Mozur

Perlez

(2017) China tech investment flying under the radar, Pentagon warns’, The New York Times. Available at: https://www.nytimes.com/2017/04/07/business/china-defense-start-ups-pentagon-technology.html (accessed 9/4/2021).

56.

Nakamura

(2014) Indigenous circuits: Navajo women and the racialization of early electronic manufacture. American Quarterly 66(4): 919–941.

57.

National Security Commission on Artificial Intelligence (2021) Final Report. Available at: https://assets.foleon.com/eu-west-2/uploads-7e3kk3/48187/nscai_full_report_digital.04d6b124173c.pdf (accessed 9/4/2021).

58.

Nishime

(2017) Whitewashing yellow futures in ex machina, cloud atlas, and advantageous: Gender, labor, and technology in sci-fi film. Journal of Asian American Studies 20(1): 29–49.

59.

Noble

(2018) Algorithms of Oppression: How Search Engines Reinforce Racism. New York: New York University Press.

60.

ÓhÉigeartaigh

Whittlestone

Liu

, et al. (2020) Overcoming barriers to cross-cultural cooperation in AI ethics and governance. Philosophy & Technology 33(4): 571–593.

61.

Roberts

Cowls

Morley

(2021) The Chinese approach to artificial intelligence: an analysis of policy, ethics, and regulation. AI and Society 36: 59–77. https://doi-org.ezp.lib.cam.ac.uk/10.1007/s00146-020-00992-2.

62.

Rodriguez

(2021) The “Asian exception” and the Scramble for Legibility: Toward an Abolitionist Approach to Anti-Asian Violence, Society and Space, April 8. Available at: https://www.societyandspace.org/articles/the-asian-exception-and-the-scramble-for-legibility-toward-an-abolitionist-approach-to-anti-asian-violence (accessed 27/09/2023).

63.

Roff

(2019) The frame problem: The AI “arms race” isn’t one. Bulletin of the Atomic Scientists 75(3): 95–98.

64.

Roh

Huang

Niu

(eds) (2015) Techno-Orientalism: Imagining Asia in Speculative Fiction, History, and Media. New Brunswick: Rutgers University Press.

65.

Roper

(2020) There’s no turning back in AI in the military, Wired, October 24. Available at: https://www.wired.com/story/opinion-theres-no-turning-back-on-ai-in-the-military/#:∼:text=Fighting%20terrorists%20for%20nearly%2020,digital%20force%20remains%20painfully%20absent (accessed 18/2/2022).

66.

Said

(1979) Orientalism. New York: Vintage Books.

67.

Said

(2001) The clash of ignorance. The Nation, October 22. Available at: https://www.thenation.com/article/archive/clash-ignorance/ (accessed 21/09/2023).

68.

Scharre

(2021) Debunking the AI arms race theory. Texas National Security Review 4(3): 121–132.

69.

Scheuerman

Pape

Hanna

(2021) Auto-essentialization: Gender in automated facial analysis as extended colonial project. Big Data & Society 8: 1–15.

70.

Schmidt

(2020) ‘Eric Schmidt: I used to run Google. Silicon Valley Could Lose to China’ The New York Times, February 27. Available at: https://www.nytimes.com/2020/02/27/opinion/eric-schmidt-ai-china.html (accessed 1/3/2022).

71.

Schuller

(2017) The Biopolitics of Feeling: Race, Sex and Science in the Nineteenth Century. Durham: Duke University Press.

72.

SCSP (n.d.) Our Mission. Special Competitive Studies Project. Available at: https://www.scsp.ai/about/what-we-do/#mission (accessed 16 January 2024).

73.

Shah

(2019) “Asians are good at math” is not a compliment: STEM success as a threat to personhood. Harvard Educational Review 89(4): 661–686.

74.

Siu

Claire

(2020) Yellow Peril and techno-Orientalism in the time of Covid-19: Racialized contagion, scientific espionage, and techno-economic warfare. Journal of Asian American Studies 23(3): 421–440.

75.

Smith

(2019) Understanding China's Confucian Edge in the Global AI Race, Quilette, 14 February. Available at: https://quillette.com/2019/02/14/understanding-chinas-confucian-edge-in-the-global-ai-race/ (accessed 18/2/2022).

76.

Sohn

(2008) Introduction: Alien/Asian: Imagining the racialized future. Melus 33(4): 5–22.

77.

Tchen

Yeats

(2014) Yellow Peril! An Archive of Anti-Asian Fear. London: Verso.

78.

Vincent

(2017) Eric Schmidt says America needs to ‘get its act together’ in AI competition with China, The Verge, November 1. Available at: https://www.theverge.com/2017/11/1/16592338/eric-schmidt-google-ai-competition-us-china (accessed 1/3/2022).

79.

Wong

(2017) Dismembered Asian/American android parts in ex machina as ‘inorganic’ critique. Transformations 29: 34–51. http://www.transformationsjournal.org/wp-content/uploads/2017/02/Transformations29_Wong.pdf .

80.

(2023) Asian Americans as “the perpetual foreigner” under scrutiny. In: Kwet

(ed.) The Cambridge Handbook of Race and Surveillance. Cambridge: Cambridge University Press, 190–222.

81.

Yang

(2016) Fiscal orientalism: China panic, the indebted citizen, and the spectacle of national debt. Journal of Asian American Studies 19(3): 375–396.

82.

Yang

(2020) Orientalism, Cyberpunk 2077, and Yellow Peril in Science Fiction’, WIRED, Dec 8. Available at: https://www.wired.com/story/orientalism-cyberpunk-2077-yellow-peril-science-fiction/ (accessed 5/12/2023).

83.

Young

(2003) The logic of masculinist protection: Reflections on the current security state. Signs: Journal of Women in Culture and Society 29(1): 1–25.

84.

Zhang

(2021) ‘Barbarising’ China in American trade war discourse: The assault on Huawei. Third World Quarterly 42(7): 1436–1454.

85.

Zwetsloot

Arnold

(2021) ‘Chinese students are not a fifth column, CSET, April 23. Available at: https://cset.georgetown.edu/article/chinese-students-are-not-a-fifth-column/ (accessed 26/09/2023).

86.

Zwetsloot

Toner

Ding

(2018) Beyond the AI arms race: America, China and the dangers of zero-sum thinking, Foreign Affairs, November 16. Available at: https://www.foreignaffairs.com/reviews/review-essay/2018-11-16/beyond-ai-arms-race (accessed 1/03/2022).

Yellow Techno-Peril: The ‘Clash of Civilizations’ and anti-Chinese racial rhetoric in the US–China AI arms race

Abstract

Keywords

Introduction

The proliferation of AI ‘arms race’ rhetoric, 2017–2023

Methodology

Results and discussion

The US–China AI arms race as a civilisational conflict

China as the ‘Yellow Techno-Peril’

Conclusion

Footnotes

Acknowledgements

Author Note

Declaration of conflicting interests

Funding

ORCID iD

Notes

References