This essay is part of a series from Carnegie’s Digital Democracy Network, a diverse group of thinkers and activists engaged in work on technology and politics. The series is produced by Carnegie’s Democracy, Conflict, and Governance Program. The full set of essays is scheduled for publication in summer 2026.

Speaking from Shanghai’s swanky World Expo Exhibition & Convention Center in July 2025, Chinese Premier Li Qiang unveiled China’s Global AI Governance Action Plan along with a proposed world AI cooperation organization (WAICO). The action plan articulated a thirteen-point road map for international AI development and governance, emphasizing multilateral collaboration on AI safety, infrastructure (such as compute, data centers, and networks), data standards, and sustainable development. The WAICO proposal, while light on logistical and administrative details, signals China’s intent to shape and sustain global AI governance institutions.

Beijing’s AI diplomacy is now pivoting from an approach focused on exporting Chinese infrastructure and associated technical standards toward a more comprehensive effort aimed at recrafting global rules, norms, and institutions of AI governance. This includes shoring up its focus on exporting and cultivating technologies that embrace Chinese ideology as well as recrafting global normative discourse on AI governance to reflect Beijing’s authoritarian interests. With Washington rapidly retreating from global cyber and AI norms-setting processes and withdrawing its financial backing for cyber diplomacy more broadly, Beijing is keen to demonstrate its global leadership. In doing so, it hopes to receive buy-in for its state-centric governance vision of technology from the Global South. 

Lessons from China’s Cyber Diplomacy

The cyber domain offers important insights into China’s global governance ambitions. China has been wary of the U.S.-brokered global internet governance regime since the emergence of the World Wide Web in the 1990s. As U.S. diplomats pursued a deregulated internet characterized by open information flows and multi-stakeholder governance, Beijing was forced to accept these norms and restrict its global engagement. Its domestic approach to internet governance—exemplified by the Great Firewall, a system designed to censor foreign online content—initially found few takers beyond China’s borders.

Today, however, information controls are proliferating across democratic and authoritarian systems alike. Authoritarian countries like Iran and Russia, along with democracies like India and Indonesia, are increasingly using tactics such as internet shutdowns, content blocking, and surveillance techniques to assert greater state control over the online information sphere. While the global proliferation of content controls primarily reflects domestic political factors and is not necessarily a blind replication of the Chinese model, these developments arguably reveal a “post-liberal” contestation of the free and open internet, providing opportunities for authoritarian countries to challenge the internet’s liberal origins.

Historically, Beijing has sought to assert influence globally by directly exporting technical infrastructure developed by Chinese companies, rather than relying on normative diplomacy. Through the Digital Silk Road initiative, China has exported a range of infrastructure, including 5G networks and facial recognition systems, to various parts of the world, including Europe, Asia, Africa, and Latin America. Chinese companies like Huawei occupy critical positions in the AI stacks for several Southeast Asian, African, and Latin American countries, enabling Beijing to exercise leverage and control over the deployment of technical infrastructure and the standards that govern them. This includes allegedly placing backdoors on infrastructure that enables China to conduct cyber espionage operations. Notwithstanding these concerns, the offerings made by Chinese companies are often cheaper and more efficient to deploy than those set forth by Western counterparts, thereby making them an attractive value proposition for the Global South.¹

China’s All-Encompassing AI Diplomacy

China views the governance of AI through the lens of entrenching state control, rather than safeguarding individual liberties. Like any leading power, China seeks to diffuse its approach globally so that the values, institutions, and standards governing AI will mirror the domestic authoritarian norms set by the Chinese Communist Party.

China’s AI strategy in this regard has hitherto been driven by infrastructural exports and an increased global market presence for Chinese companies. However, this is complemented today with a more holistic diplomatic strategy that also seeks to reshape normative rhetoric at international forums. China’s aim is to influence AI global governance in a manner that suits its own interests—much like the United States crafted the liberal internet governance regime to reflect democratic values as well as to benefit Western companies.

In this regard, Beijing is amplifying its influence not only within UN institutions, but also through informal bodies that complement UN processes. In 2024, for example, the UN General Assembly adopted by consensus a Chinese-sponsored resolution connecting AI, sustainable development, and national sovereignty. This resolution was tabled by a core Group of Friends for International Cooperation on AI Capacity-Building comprising eighty countries and UN entities. Co-chaired by China and Zambia, the group hosts events and seminars to further cooperation and dialogue on AI capacity-building. Moreover, China’s 2025 Global AI Governance Action Plan acknowledges the UN as the “main channel” for crafting an equitable global governance framework and explicitly references the Global Digital Compact, a comprehensive UN-brokered road map for global digital cooperation. China’s support for these processes is a clear sign of its interest in advancing international AI dialogue under the auspices of the UN, a body that it increasingly believes it can control.

China’s WAICO, when operationalized, will likely complement its efforts at the UN. Although its functional and operational details remain unclear, it is likely that China will use it to build a preliminary consensus on AI governance issues and subsequently leverage this consensus to coordinate positions at the United Nations. While global AI governance discourse is in its nascent stages, China has laid the groundwork to shape future discussions in line with its preferred authoritarian approach at home.

In addition to its normative push through the UN and informal institutions, China has also focused on building technical and governance capacity in developing countries through workshops and seminars. It will likely use WAICO to expand and consolidate its network of AI cooperation centres, including the China-BRICS AI Development & Cooperation Centre (set up in January 2025 and headquartered in Shanghai’s Xuhui district) and the China-Laos AI Innovation Cooperation Center, which are designed to facilitate building AI infrastructure and cultivating AI talent.

Such capacity-building efforts certainly offer important developmental opportunities for partnering countries. However, Beijing would likely continue to assert control over the operation of the technologies underpinning AI. For example, any large language model made by a China-based company might automatically censor content embracing liberal norms such as the right to protest, regardless of where the model was deployed.

In essence, Beijing’s efforts to shape multilateral rhetoric on AI at the United Nations, while simultaneously pursuing technology diffusion and capacity-building abroad, represent two sides of the same coin. The latter expands China’s material power while the former helps legitimize its state-centric approach to AI governance.

What Now Under Trump?

China’s confidence in its AI diplomacy today is driven both by its increasing AI capabilities as well as the belief that its development-oriented vision is likely to be more appealing to the Global South.

Beijing’s diplomatic ascent is reinforced by America’s AI policy reversal under President Donald Trump. As China pivots toward norms-oriented diplomacy, the United States seems to be embracing economic muscle-flexing and diplomatic coercion.

At the Paris AI Summit in February 2025, Vice President JD Vance made it clear that the U.S. priority was to win the AI race and that concerns about safety were hindrances. One year later at New Delhi’s AI Summit, the U.S. delegation emphasized promoting American “dominance” of the AI stack. “We want to make sure the world uses the American AI stack. That means from the bottom up,” said White House AI Policy Advisor Sriram Krishnan. “It means our semiconductors and advanced AI chips, Nvidia, AMD, Google’s TPUs, and so on and so forth.” 

In contrast to WAICO’s focus on multilateralism and development, America’s AI Action Plan under the Trump administration prioritizes accelerating innovation, building AI infrastructure domestically, and promoting U.S. standards through its infrastructure. To the extent that the plan touches on diplomacy, it is limited to “counter[ing] Chinese influence in international governance bodies” rather than articulating an agenda of its own.

The changing U.S. approach to AI governance is accompanied by a hollowing out of the State Department’s cyber diplomacy institutions through budget and personnel cuts. State Department–funded civil society institutions that were working at multilateral organizations to counter China’s diplomatic push on norms and standards are also constrained. The Trump administration marked the new year by pulling out of several formal and informal institutions, including notably the Freedom Online Coalition and the Global Forum on Cyber Expertise, that focused on promoting digital human rights and crafting cyber capacity-building, respectively.

In contrast, China-backed private sector enterprises, academics, and so-called government-organized nongovernmental organizations are filling the void. With Washington retreating from its commitment toward a free and open internet and retrenching from global AI institutions, engaging in norms formulation and capacity-building provides an opening for Beijing to hollow out liberal norms as they apply to the deployment of AI stacks in the Global South.

Will China Win the Global South? 

Beijing’s AI vision has yet to win wholesale buy-in from the Global South. Developing countries have not explicitly endorsed Beijing’s diplomatic push on AI nor have they backed the new institutional architecture proposed by Beijing, even as they selectively use Chinese AI products or participate in informal seminars and forums on AI capacity-building. The impact of China’s and Russia’s cyber diplomacy over the past decade suggests that wholesale, normative alignment with their state-controlled vision is unlikely.

Although developing countries are wary of ideological spats, they remain open to engaging with different leading powers including China, the EU, and the United States to further their own digital development goals. While many countries in the Global South are willing to consider China’s state-centric and public order–oriented vision of AI, they also see value in governing AI through liberal norms and human rights safeguards.

Multilateral processes on AI governance, including the Global Dialogue on Artificial Intelligence, are taking shape at the UN, providing a platform for Global South countries to participate in and shape global norms-setting processes. The developing world retains an opportunity to articulate norms that prioritize development and AI capacity-building while respecting the rights of individuals and communities impacted by AI. Despite differing interests and approaches to international engagement, developing countries should not be forced to choose between China’s diplomatic expansion or the Trump administration’s unpredictable mercantilism. For many, a path rooted in democracy and the rule of law remains the most prudent way forward.

Acknowledgments

I am thankful to Steve Feldstein, Shreya Joshi, Valentin Weber, and members of the DDN for feedback on previous drafts. All errors remain my own.