Just as space dominance once symbolized global power during the Cold War, artificial intelligence now sits at the center of a high-stakes geo-strategic competition between the U.S. and China. But technological breakthroughs alone will not determine the outcome. A parallel battle over governance, in terms of who sets the rules, norms, and standards, may prove equally decisive in shaping not just AI, but the global balance of power.

In late July, China announced its Global Artificial Intelligence Governance Action Plan. The timing was conspicuously close to the Trump administration’s own recently released strategy, entitled Winning the Race. While Western attention has often focused on Beijing’s AI surveillance state and its drive for frontier-model supremacy, this plan signals a notable rhetorical, if not strategic, pivot. Increasingly, China now casts itself as a responsible global actor in contrast to a more isolationist and combative United States. By championing multilateralism and “shared governance” in AI, Beijing aims to embed its technologies and influence abroad before Washington can set the agenda. As Liza Tobin argued in her analysis of Xi’s global governance strategy, these moves are rarely ad hoc in nature. Rather, they form part of a systematic effort to reshape international norms and institutions in China’s favor. The AI plan should be read as a continuation of this broader project.

The US plan frames artificial intelligence as a transformative technology capable of shifting the global balance of power. It positions AI development and governance as a national security imperative, integrating domestic innovation, strategic partnerships, and foreign policy to safeguard American interests. A central pillar of the strategy is building a US-led alliance, promoting US-aligned safety standards, software, and hardware, and creating a web of interoperability that binds allies and partners to the American technology stack.

Beijing’s version contrasts sharply. It focuses solely on bilateral and multilateral cooperation; a somewhat awkward departure from its long-standing doctrine of “cyber sovereignty.” Critics argue China’s pledge to “facilitate the lawful, orderly, and free flow of data” is less about liberalizing information exchange and more about projecting openness while retaining tight state control. However, China’s shift is not a reversal but an evolution of the pattern Tobin identified, packaging sovereignty-first instincts in the rhetoric of openness and inclusivity in order to normalize them globally. It is also a clear attempt to counter the US and UK narratives that AI governance and development is not just a race, but a civilizational contest between democracies and authoritarian states.

This follows the unexpected domestic success of DeepSeek’s R1 model, which boosted Beijing’s confidence in its ability to compete at the frontier of development, despite US semiconductor restrictions. DeepSeek mattered for more than just symbolic catch-up. Developed on Chinese-made AI accelerators and GPUs, including domestically produced Hygon and Ascend chips, and optimized for efficiency and lower costs, R1 hit performance benchmarks competitive with leading US models while relying on fewer high-end Western chips, blunting the impact of Washington’s export controls. Its release was heralded in the language of the Cold War arms race as a Sputnik moment, demonstrating that China could innovate around chokepoints in the AI supply chain. DeepSeek turned what had been seen as China’s vulnerability into a proof-of-concept for self-reliance in the AI domain. The technical leap has emboldened Chinese policymakers and tech leaders to speak more confidently about China’s role in shaping global AI rules. Indeed Premier Li Qiang launched the plan at a Shanghai conference attended by global tech leaders, investors, and policymakers, framing China as a consensus-builder amid escalating US-China tensions. Beijing also advocated for the creation of a new international AI cooperation body, headquartered in Shanghai, to complement the UN’s work.

A strong role for the UN is central to China’s competitive calculus and institutional strategy. Beijing has previously used the institution as a brokering partner for economic development projects. China seeks not just to participate in multilateral institutions but to transform them into vehicles for embedding Chinese preferences. As Kristen Cordell and Daniel Kliman have documented, Beijing has also invested heavily in standards-setting bodies like the ITU, leveraging these technical forums to advance Chinese frameworks and establish first-mover advantage in rule-writing. Standards are the invisible architecture of globalization, and by shaping them early, Beijing aims to ensure its technologies are “locked in” worldwide, making it difficult for competitors to displace them later.

Washington’s approach is rooted in standards-setting as a national strategic instrument. Rather than pursuing binding global regulations, it seeks to embed US norms, emphasizing openness, innovation, and intellectual property protection, through bodies like the ISO, ITU, and OECD. This “soft infrastructure” strategy aims to maintain global dependence on US technologies without requiring universal legal consensus or norms. Its success will hinge on whether US technology remains both superior and attractive to Chinese alternatives, and whether allies view the US framework as promoting shared interests rather than simply strategic advantage. As Washington leans on innovation and alliances, Beijing doubles down on institutional positioning. This divergence raises the possibility that even if US firms lead commercially, China could still win politically by defining the rules others must follow.

China’s plan characterizes AI as an “international public good,” a framing designed to appeal to emerging economies wary of US techno-hegemony. By promoting equitable access, open-source collaboration, and inclusive governance, Beijing positions itself as a leader aligned with the development needs of the Global South. Again, this reflects a broader governance strategy. China appeals to shared prosperity while creating long-term structural dependencies on its technology and institutions. Beijing has consistently projected openness abroad while retaining strict control at home, and has sought to export this hybrid model through infrastructure and standards. As with the Belt and Road Initiative, this involves exporting AI-ready infrastructure, from 5G/6G networks, to data centers, and intelligent computing hubs, embedding China’s technology stacks and service contracts in partner nations, thereby creating enduring political and economic dependencies. For many states in Africa, Southeast Asia, and Latin America, the attraction is not ideological but highly practical. Chinese infrastructure includes not just financing, but turnkey systems, and governance templates. This makes Beijing’s AI governance pitch especially persuasive in regions where governments see technology as a development lever rather than a security risk.

The two approaches echo Cold War-era diplomacy, when both the United States and USSR claimed to champion peace while competing for technological dominance. Yet arguably AI governance is a more complex challenge than nuclear arms control ever was, since it is deeply intertwined with trade, market interdependence, and civilian applications. Unlike the “Pax Americana” of the mid-twentieth century, today’s geopolitical environment lacks a dominant guarantor of the multilateral order. In this way, AI’s fusion of economic competition and national security makes governance both more necessary and more politically complex.

Both sides could, in theory, cooperate on narrow issues such as preventing non-state actors from acquiring high-risk AI capabilities. But the broader contest over norms, standards, and the infrastructure of the AI age will define not just the technology’s trajectory, but the global balance of power in the decades ahead. In this sense, the AI governance race represents a new domain of the same transformational agenda. Beijing’s major effort is to write rules for a world order where China, not the U.S., sets the terms. Whether Washington can respond effectively will depend less on outspending Beijing and more on outmaneuvering it in the internationally regulatory arena. By ensuring allies see American frameworks not simply as instruments of US hegemony, but as genuine public goods.