US-China AI rivalry needn’t preclude cooperation

AI technology presents global risks that transcend national boundaries, encompassing issues such as cyberattacks, misinformation, system failures, and potential misuse. In an increasingly geopolitical landscape, does a lack of compatibility between nations inherently enhance AI safety? At the inaugural UN Global Dialogue on AI Governance, UN Secretary General Antonio Guterres provided a definitive response: "When countries agree on how to test systems, assess risks, and assign responsibility, safety travels with the technology."

Conversely, "a patchwork of incompatible rules raises costs, divides the world – and protects no one."

Despite the apparent need for AI-driven cooperation, geopolitical tensions pose a significant threat to global unity. In response, China launched the World AI Cooperation Organization (Waico) alongside 29 founding members, focusing on AI cooperation and governance, as well as assisting developing nations in closing the AI gap. Meanwhile, the United States developed the Pax Silica initiative, emphasizing AI, technology, and supply-chain security.

While both initiatives share certain objectives, they primarily address distinct concerns. Waico, originating from China, seeks to foster development and cooperation while expanding China's global AI governance influence; Pax Silica, emerging from the United States, aims to tackle genuine critical technology and supply-chain resilience concerns while cultivating a network of US-aligned partners.

The presence of these rival initiatives raises an analytical challenge when geopolitics is perceived as the sole determinant of their significance. Indonesia joined Waico while simultaneously participating in Pax Silica discussions, and Kazakhstan has also joined both. The United States is reportedly contemplating a policy that would exclude countries from engaging in initiatives deemed incompatible with its framework.

This raises a critical question: must international AI cooperation inadvertently replicate geopolitical blocs? Geopolitical conflicts stem from diverse commercial, strategic, and political interests, but the way these conflicts are perceived matters significantly. Once a nation is primarily viewed as a strategic adversary, cooperation might be interpreted as bloc-building, regulation as political control, and ambiguous actions as evidence of hostile intent.

This is not merely a theoretical concern. A Carnegie Endowment for International Peace paper warns against dismissing Chinese AI regulations due to the Communist Party's dominance or viewing them through a purely geopolitical lens.

The complexity lies in the fact that political and information control are important factors, but policymakers are also driven by the social, economic, and ethical implications of AI. Geopolitical differences can simultaneously coexist with areas of policy convergence. A 2025 study examining over 40 policy and corporate governance documents from the US and China identified common ground in system reliability, algorithmic transparency, and the use of AI for safety enhancement.

The error does not lie in acknowledging political motives but in assuming that political motives are the sole explanations. For instance, Chinese President Xi Jinping advocated for human control of AI during the World AI Conference in Shanghai, while the UN Global Dialogue on AI Governance emphasized robust human oversight among its governance priorities.

While the language used by both entities may appear similar, it does not necessarily imply identical motivations, objectives, or implementation methods. However, institutional differences do not automatically signify fundamentally divergent policy concerns.

To gain a comprehensive understanding, one must examine the specific issues a policy aims to resolve and how it functions within its institutional context. A country's political system should be comprehended within the broader framework of its history and culture, rather than being judged against predetermined assumptions about what the system ought to be, which could inadvertently pre-determine the policy's meaning.

As China's political landscape is heavily influenced by the Communist Party, recognizing this aspect is crucial but should not be used as an all-encompassing explanation. An AI policy may reflect political considerations, technological safety, economic development, national security, or even a combination of these objectives simultaneously.

The choices made by countries like Indonesia and Kazakhstan suggest an alternative approach: participation without alignment. Countries can engage within different technological ecosystems while maintaining the flexibility to implement independent policies. Forcing them to choose between two competing frameworks may exacerbate the very divisions that global AI governance seeks to bridge.

The concept of "he er bu tong" – harmony without uniformity – aptly captures this possibility. Countries shaped by different histories, cultures, and circumstances need not share political systems, values, or strategic interests to cooperate effectively. As AI technology continues to evolve, it is imperative for global leaders and analysts to cooperate across differences.

They must acknowledge genuine differences while also recognizing the common ground that exists. The responsibility falls on journalists, think tanks, and academics to accurately identify both disparities and shared objectives, distinguishing facts from inferences. By taking these steps, AI can serve humanity rather than deepen geopolitical divisions.