Executive Summary

Two tectonic trends in the international security environment appear to be on a collision course. The first trend is the intensifying geopolitical rivalry between the United States and the People’s Republic of China (PRC or China). The second trend is the rapid development of artificial intelligence (AI) technologies, including for military applications. This report explores how the United States can manage strategic risks—defined as increased risks of armed conflict or the threat of nuclear war—that could be created or exacerbated by military AI in its relationship with China.

It begins by providing an overview of China’s views on and policies toward AI. Beijing sees AI playing roles in both its civilian economy and the modernization of its military, the People’s Liberation Army (PLA). At home, Chinese leaders want to leverage AI to boost growth and innovation, address economic and social challenges, and secure the Chinese Communist Party’s (CCP) domestic rule.

AI also plays a key role in China’s military ambitions, especially its goal to become a “world-class military” by midcentury, in part through the “intelligentization” of its forces. Intelligentization relies on integrating AI and other emerging technologies into the joint force with the goal of gaining an edge on the United States. China argues that its governance model, including its military-civil fusion policy, gives Beijing a competitive advantage over Washington. Realization of that vision, however, remains uncertain and will require China to overcome external and internal obstacles.

This report explores how the United States can manage strategic risks—defined as increased risks of armed conflict or the threat of nuclear war—that could be created or exacerbated by military AI in its relationship with China.

Next, the report articulates five categories of what the authors call pathways, or causal links, through which applications of military AI could undermine stability and increase strategic risk between Washington and Beijing. The first is individual improvements in capabilities that combine to give China a military edge. The second is AI’s effects on the decision-making and information domain. The third is uncrewed autonomous systems. The fourth is intelligence, surveillance, and reconnaissance. And the fifth is command, control, and communications. The report’s discussion of each pathway provides more details about the intricacies of how they might function.

Taken together, the emergence of military AI will likely deepen U.S.-China rivalry and increase strategic risks. Policymakers can draw on three categories of response options. These include limiting China’s military AI while advancing the United States’ own capabilities; engaging in unilateral responsible management of military AI; and pursuing bilateral and multilateral diplomacy to reduce strategic risks. Washington has already taken actions across all three categories, and foreign capitals have likewise sought to shape the military AI environment. To build on the actions taken to date, U.S. policymakers should:

Take bold action to constrain China’s progress in AI for military and repressive purposes, but do so in a narrow way that avoids self-defeating steps;
Build U.S. military AI capabilities to stay on the cutting edge;
Develop, promulgate, and implement norms and best practices on responsible military AI;
Proactively engage with like-minded allies and partners and in multilateral institutions on military AI issues;
Negotiate risk reduction and confidence-building measures with China related to military AI;
Continue to pursue universal U.S.-China risk reduction and crisis management mechanisms despite persistent challenges;
Make military AI a fundamental pillar of diplomacy with China related to nuclear weapons and strategic stability;
Take steps to reduce strategic risks not directly caused by military AI but potentially worsened by the inherent speed and unpredictability of military AI;
Prioritize intelligence-gathering and analysis on, and net assessment of, China’s military AI capabilities.

Introduction

Two tectonic trends in the international security environment appear to be on a collision course. The first trend is the intensifying geopolitical rivalry between the United States and the People’s Republic of China (PRC or China). Since the mid-2010s, relations between Washington and Beijing have descended into what U.S. President Joe Biden has called “extreme” strategic competition that spans the diplomatic, economic, military, and governance arenas.¹ Leaders from both countries have further asserted that emerging technologies will play a decisive role in the contest between the two superpowers over the future of global power, including by shaping the military balance.

The second trend is the rapid development of artificial intelligence (AI) technologies. Recent AI breakthroughs, such as ChatGPT, DALL•E, and other generative AI models, have splashed across headlines worldwide and racked up astronomical user numbers in record time.² They illustrate how emerging technologies might revolutionize entire sectors quickly once they reach maturity. Major developments in AI have largely been confined to the civilian sector so far. However, AI systems that can be used for a range of military applications—a broad category this report refers to using the shorthand military AI—are likewise evolving swiftly. The combination of these two trends has the potential to increase what this report calls strategic risks. That term is used here to refer to increased risks of armed conflict or the threat of nuclear war between Washington and Beijing.³

This report explores how the United States can manage strategic risks that could be created or exacerbated by military AI in its relationship with China. The report proceeds in four sections. The first section details how China sees AI playing roles in both its civilian economy and the modernization of its military, the People’s Liberation Army (PLA). It also explores how Beijing sees these emerging capabilities shaping the future character of warfare.⁴ The second section details five categories of what the authors call pathways, or causal links, through which applications of military AI could undermine stability and increase strategic risk between Washington and Beijing. The third section assesses three categories of options for managing strategic risks along those pathways. Finally, the fourth section offers recommendations for how U.S. policymakers can reduce and mitigate the risks military AI might pose to peace and stability in strategic competition with China.