China as an analytical lens for AI and society

Abstract

How can “China” serve as a productive analytical tool for understanding the relationship between artificial intelligence (AI) and society? This commentary proposes a conceptual typology to answer that question. We offer four ways to use China as an analytical lens—China as Mirror, China as Prototype, China as Counterpoint, and China as Global Actor. Each concept illustrates how China's particularities and global entanglements can help reframe core sociotechnical questions about AI's political economy, labor, subject formation, ideology, and global circulation. This typology contributes to the decolonization of knowledge production in critical AI studies, ensuring that insights are co-generated globally and avoiding essentialized boundaries that hinder this effort.

Introduction

In public debates on artificial intelligence (AI), China often appears as a specter. At times, it is a technological powerhouse racing for supremacy; at other times, it is a high-tech dystopia exemplifying AI-driven authoritarianism. This commentary asks what becomes possible when we treat “China” not as an object of alarm or exception, but as an analytical lens for understanding AI and society—a vantage that lets us see how theories travel, where they falter, and how they must be revised.

Current research in dynamics of AI and society has largely been rooted in West, carrying implicit assumptions about how capitalism drives AI, what appropriate governance looks like, or how individuals relate to AI systems. Using China as an analytical lens decenters these assumptions. It echoes Chen's (2010) call to use Asia as method for deimperializing knowledge, by developing “multiple reference points” that provincialize the West's self-image as the universal norm.

We outline four approaches for using “China” to enrich our understanding of AI and society: (1) China as Prototype considers China as a testing ground for large-scale AI integration and governance innovations, offering a glimpse of potential futures and alternatives. (2) China as Mirror highlights how China's AI realities reflect global challenges forcing Western observers to recognize similar yet ignored dynamics at home. (3) China as Counterpoint treats the Chinese experience as a comparative foil that challenges “universalist” theories. (4) China as Global Actor emphasizes China's role in the global circulation of AI technologies and discourses. Note that these are not mutually exclusive categories, but rather analytical angles that shed light on different facets of the issue.

We aim to leverage China's particularities and global engagements to critique Western centrism without falling into the trap of Chinese exceptionalism (Franceschini and Loubere, 2022), where specificities become exotic details reinforcing epistemic boundaries between Global South and North (Go, 2023). Instead, by analyzing AI through China's unique position in global networks, we expose the limits of supposedly universal models when confronted with different social settings.

China as prototype

One productive analytical approach positions China as a prototype. It sees China as a frontrunner of state-led AI development and social integration, offering insights into potential futures and alternative models of sociotechnical configurations. This perspective invites us to study Chinese implementations as early demonstrations of possibilities that may emerge elsewhere.

China demonstrates an alternative political economy for developing AI industries. Contrasting to the model in the United States that prioritizes domestic private sector innovation without heavy-handed state intervention or in many other countries that relies on international tech giants (Kotliar and Gekker, 2024), the Chinese state never conceals its capacity to intervene in the private sector and align it with state interests (Lei, 2023; Lindtner, 2020).

This dual approach operates in complementary ways. On the one hand, the government designs industrial policies to encourage AI development in strategic areas through selected “national champions,” intertwining technological advancement with nationalist narratives (Roberts et al., 2020). For instance, Alibaba—originally a retail and e-commerce company that has since expanded into cloud service, fintech, and entertainment—was one of the selected national champions and was entrusted with the development of AI for smart cities. The much-discussed “City Brain” project in Hangzhou, run by Alibaba, is often cited as a successful prototype of AI-enhanced urban management (Cai, 2025).

On the other hand, the Chinese state utilized its growing capacity to strengthen its control over the private sector. In 2021, Alibaba paid a record-high penalty of $2.8 billion to the antimonopoly authority for allegedly abusing its market position. Indeed, state governance in China tends to swing cyclically between the deployment of arbitrary power, characterized by campaign-style governance and crackdown, and routine-based power that relies on its bureaucracy (Lei, 2023).

While such state-market interventions were long considered uniquely Chinese, recent developments in the United States and Europe—from emerging “AI nationalism” discourses (Kak and West, 2024) to the Trump administration's 2025 AI Action Plan (U.S. White House, 2025)—suggest that China could rather be understood as a prototype rather than an outlier. This does not imply technological determinism where all societies follow China's path; rather, it encourages analyzing varied implementation pathways shaped by different political economies (Lindtner, 2020). China as prototype means using China's expansive and diverse AI configurations and implementations as a forward-looking case study. This perspective encourages researchers to move beyond abstract speculation about AI futures by analyzing empirical evidence from China's present. It expands the scope of questions about AI's social impacts: we ask not only “What if … ?” but “What has happened when … ?” at a scale of hundreds of millions of people.

China as mirror

Another analytical approach is to use China as a mirror, looking closely at China's AI landscape reflects patterns that can be also observed elsewhere—sometimes more sharply—so that parallels, not just contrasts, come into view. This perspective draw attention to how social dynamics, power struggles, and regulatory mechanisms around AI in China illuminate similar processes elsewhere that might otherwise remain overlooked or obscured.

At first glance, the prominent role of the Chinese state in making industrial policies and selecting national champions presents a sharp contrast to the US market-centered approach. A longer historical view, however, reveals important similarities. Even though the US government did not formally articulate industrial policies, it has played a crucial infrastructural role in supporting AI development through military funding, federal research grants, and enabling regulatory frameworks (Farrell and Newman, 2023). The mirror reflects uncomfortable parallels that Western observers may be reluctant to acknowledge.

Indeed, China shares more similarities with Western nations than popularly assumed. Analysis of policy documents shows that Chinese government narratives about AI, which present it as both inevitable and dangerously uncertain, mirror rhetoric from the United States, France, and Germany (Bareis and Katzenbach, 2022). These parallels invite critical examination of Western AI governance. For example, research on Chinese AI surveillance encourages scholars to reconsider similar systems in Western societies, such as predictive policing algorithms disproportionately targeting minority communities (Burrell and Fourcade, 2021; Franceschini and Loubere, 2022). Rather than dismissing Chinese surveillance as exceptional, the mirror reveals concentrated versions of broader global trends.

Seeing the similarities is especially critical amid current geopolitical tension, where “Chinese AI” has increasingly taken on a life of its own as a discursive object—detached from the realities of AI development in China, as the recent case of DeepSeek shows (Yuan, 2025). The perceived differences between “Chinese AI” and “US AI” have been strategically weaponized by both Chinese and US stakeholders to serve divergent interests (Tan and Weigel, 2022; von Blomberg and Liu, 2025). Under such tension, similar problems with AI in the United States, such as AI discrimination, may escape public criticism by essentializing the differences between China and the West (McInerney, 2024). Consequently, “Chinese AI,” much like earlier “Chinese Internet” or “Chinese innovation,” risks becoming Techno-Orientalized, “a free-floating signifier for the omnipotent digital dystopia, without history, culture, or institutional complexity” (Chen et al., 2023: 259).

In sum, treating China as a mirror means using the Chinese case to reflect common challenges faced by global AI and reveal Western blind spots that have been intentionally ignored. It reveals that phenomena like government intervention, mass surveillance, and AI-driven ideology are not “China problems” but global issues that manifest in diverse forms, fostering critical self-awareness.

China as a counterpoint

The third analytic move is to treat China as a counterpoint, not only to highlight how its case differs from other contexts, but to leverage these differences to rethink and revise prevailing theories of AI and society. China's blend of socialist legacy and capitalist reform, its authoritarian governance, and its distinct cultural frameworks provides a powerful counterpoint to the liberal-democratic, late-capitalist contexts that underpin much of Western AI scholarship. Using China as a counterpoint thus prompts the question: How must our prevailing theories about AI and society be re-examined when situated in a markedly different sociopolitical context?

Consider the data-annotation industry that has underwritten recent AI advances. In China, its organization diverges sharply from US and European models: rather than outsourcing data annotation overseas, Chinese tech companies assign such work to less developed domestic regions, often in collaboration with local governments (Wu et al., 2025). As a counterpoint, the Chinese model recasts annotation from dispersed global gig work to locally embedded, state-coordinated labor, unsettling canonical accounts of platform capitalism and data colonialism.

Subject formation and citizen experience under AI likewise take distinct shape. While “coding elites” (Burrell and Fourcade, 2021) in China also enjoy material and symbolic privileges, their power is constrained compared to their US counterparts, and they often experience intense overwork in the “996” culture (9am–9pm, 6 days weekly) (Lei, 2023; Tan and Weigel, 2022). These contrasts underscore how differing working cultures, organizational structures, and economic systems shape AI subjects and channel the strategies and agendas of labor movements. More striking counterpoints emerge in citizen responses to surveillance technologies. Broad public acceptance of securitization discourses in China helps legitimate surveillance infrastructures (Liu and Graham, 2021), complicating liberal assumptions about universal privacy norms and revealing how subjective experiences of AI systems are mediated by cultural narratives and perceived benefits of social order.

Using China as a counterpoint therefore challenges Western-centric theory building and taken-for-granted assumptions, adding comparative leverage to the studies of AI's social impact. The aim is not to swap US exceptionalism for Chinese exceptionalism, nor to replace technological determinism with political or cultural determinism that risks Orientalizing China to justify the AI “arm race” discourse (McInerney, 2024; von Blomberg and Liu, 2025). Rather, it asks scholars to revisit broad claims about AI's social impact and consider contextually grounded analyses. In the end, this approach uses China not as an outlier to be explained away, but as a theory-building case that expands the conceptual vocabulary.

China as a global actor

The final conceptual lens situates China within the global networks of technology, capital, and ideas—viewing China as a global actor in the AI arena. This perspective resonates to recent call to adopt “global China” as am method (Franceschini and Loubere, 2022): rather than treating China as a self-contained case, it foregrounds global entanglements—how China both shapes and is shaped by international flows of AI products, talent, standards, and discourse.

China's position in global AI networks manifests through multiple channels. Materially, rare earth minerals mined in China are crucial for the production of AI hardware. Many of these hardware are also manufactured in Chinese factories that depend on Western-designed and globally produced semiconductors (Shih, 2021). The global flow of people, ideas, and patents that connect China and the rest of the world is also not an abstraction but how research, production, and implementation of AI systems. Meanwhile, Chinese companies like Huawei and ByteDance compete actively in the global market in exporting AI products beyond its boarder, prioritizing investing in markets that are overlooked by US tech giants (Yakefu, 2024).

While earlier neoliberal globalization frameworks obscured contested networks in AI development, recent trade restrictions on AI-related capital, technology, and market access have forced China to develop more distinct approaches to AI research and implementation (Chen et al., 2023; Jiang, 2024). For example, the US embargo on Nvidia's high-end chips pushed DeepSeek to turn to domestic alternatives such as Huawei (Yuan, 2025). Taking China's role in these networks seriously, therefore, allows scholars to move past nation-centric analysis toward a network analysis of AI's social impacts, seeing China not as an island, but an integral player shaping the global AI landscape.

This also means to see China as one node in the global network of AI, thus moving beyond the US–China rivalry that has been dominating the global discussion. Nations like Brazil, India, and South Africa increasingly develop independent AI strategies and regulatory frameworks that draw from multiple influences rather than following either Chinese or American models (Jiang, 2024). Especially, it is critical to consider the agency of Global South countries in AI's development, regulation, and implementation, and to seek theoretical traditions from their sociohistorical contexts—traditions that were erased or made invisible—to rethink the contemporary concerns (Valente and Grohmann, 2024). The global actor lens thus reinforces a key message: AI's social impacts cannot be comprehended through a single-nation story or a bipolar cold-war story. China, despite its importance, is only one of the threads interlacing with others in the network.

Conclusion

This commentary has proposed a typology of four analytical lenses—China as Mirror, Prototype, Counterpoint, and Global Actor—to illustrate how “China” can be used as a productive analytical tool for rethinking AI's social contexts and impacts. Together, these perspectives demonstrate the theoretical richness that emerges when we take China's experience seriously without exoticizing it. By demonstrating how China can serve as an analytical lens, we underscore the importance of contextualizing AI in global and comparative perspective. It is also in line with broader moves in social science to decolonize knowledge production, ensuring that insights are co-generated from different parts of the world rather than imposed from a single center (Chen, 2010; Chen et al., 2023; Go, 2023).

For researchers applying these lenses, several methodological implications emerge. The Prototype lens benefits from longitudinal studies tracking how specific AI implementations evolve over time in Chinese contexts before appearing elsewhere. The mirror approach requires symmetrical analysis comparing similar technologies across different societies using consistent analytical frameworks. Counterpoint analysis works best with careful case selection identifying comparable technological applications in contrasting social settings. Finally, the Global Actor perspective demands multisited research following technologies as they circulate through global networks.

In approaching “China and AI” in this way, it is essential to maintain a critical balance. The goal is neither to vilify nor glorify China, but to learn from it. We highlight China's specificities, which, as Irani et al. (2010) argued, are “not a problem to be solved, but a reality that should be central” to research. Our perspective further requires us not only to incorporate China into the research of AI as a unified object in comparison, but also to recognize China's internal complexities, its sociotechnical contexts, and its connections with other societies (Chen et al., 2023; Franceschini and Loubere, 2022). By engaging with those particularities, we equip ourselves to ask better questions about AI in any society. Otherwise, the emphasis on non-Western knowledge production risks inadvertently reinforcing essentialized boundaries between societies that undermine the decolonization endeavor. Also, we note that these lenses are not static; they may evolve as AI technologies, governance regimes, and global dynamics shift, underscoring the need for continued reflexivity in their application.

Eventually, this conceptual typology contributes to building a more pluralistic and reflexive analytical framework for examining AI's global sociotechnical configurations and impacts. By decentering Western experiences without exoticizing alternatives, these lenses help scholars navigate between technological determinism and cultural essentialism, revealing how AI technologies emerge from and reshape varied social contexts. As AI continues impacting societies worldwide, such approaches will be increasingly vital.