The People’s Liberation Army is accelerating its investment in artificial intelligence-enabled unmanned systems and swarming technology, drawing lessons from the conflicts in Ukraine and the Middle East to reshape its military doctrine and acquisition priorities. A new analysis published by the American Enterprise Institute on April 7 examines the PLA’s efforts to adapt to contemporary warfare, finding that while the Chinese military has made significant progress in integrating AI into its unmanned systems, it continues to face institutional and cultural constraints that limit the speed and depth of its adaptation. The report provides one of the most detailed assessments to date of how China is translating its civilian AI capabilities into military applications, and what the implications are for US defense planning.
Lessons from Ukraine and the Middle East
The AEI analysis identifies the conflicts in Ukraine and the Middle East as the primary drivers of the PLA’s current doctrinal evolution. Both conflicts have demonstrated the decisive importance of unmanned aerial systems, from low-cost commercial drones modified for military use to sophisticated loitering munitions, and have provided a real-world laboratory for testing AI-enabled capabilities including autonomous target recognition, swarm coordination, and electronic warfare. Chinese military planners have been studying these conflicts intensively, and the lessons they have drawn are reflected in the PLA’s recent procurement decisions and training exercises.
The most significant lesson, according to the AEI analysis, is the importance of mass. The conflicts in Ukraine and the Middle East have shown that the ability to deploy large numbers of relatively inexpensive unmanned systems can overwhelm more sophisticated but numerically inferior defenses. This finding aligns well with China’s existing strengths, its manufacturing capacity, its large pool of AI engineering talent, and its experience deploying AI systems at scale in civilian applications. This has reinforced the PLA’s commitment to developing AI-enabled swarming capabilities that can saturate enemy air defenses with coordinated attacks by hundreds or thousands of autonomous systems.
The PLA’s AI Integration: Progress and Constraints
The AEI report acknowledges that the PLA has made substantial progress in integrating AI into its unmanned systems over the past several years. Chinese military drones are increasingly equipped with AI-powered target recognition systems, autonomous navigation capabilities, and the ability to operate in GPS-denied environments. The PLA has also conducted a series of exercises that have tested swarming tactics at increasing scale, demonstrating the ability to coordinate the actions of large numbers of unmanned systems in complex operational environments.
However, the report also identifies significant constraints on the PLA’s ability to fully exploit its AI capabilities. Chief among these is an institutional culture that prioritizes control and predictability over the kind of autonomous decision-making that AI-enabled systems require to be most effective. PLA doctrine has historically emphasized centralized command and control, and the shift to a model in which AI systems make consequential decisions autonomously, including potentially lethal ones, requires a fundamental change in how the military thinks about the relationship between human commanders and machine systems. This cultural shift is occurring, but it is occurring more slowly than the pace of technological development would allow.
Implications for US Defense Planning and the Taiwan Strait
The PLA’s accelerating investment in AI-enabled unmanned systems has direct implications for US defense planning, particularly with respect to the Taiwan Strait scenario that dominates US military planning in the Indo-Pacific. A large-scale Chinese military operation against Taiwan would almost certainly involve the extensive use of AI-enabled drones and swarming systems, both to overwhelm Taiwan’s air defenses and to complicate US naval operations in the region. The AEI analysis suggests that the PLA’s capabilities in this area are advancing faster than many US defense planners have assumed, and that the window for the United States to develop effective countermeasures may be narrowing.
For the broader US-China AI competition, the PLA’s military AI program is a reminder that the stakes of the competition extend well beyond economic competitiveness and technological leadership. The integration of AI into military systems is creating new forms of strategic risk that neither country fully understands, and that both countries have an interest in managing carefully. As the Foreign Affairs essay published earlier this week argues, this shared interest in avoiding catastrophic outcomes may ultimately provide the foundation for a degree of US-China cooperation on AI safety — even in the context of an intense and ongoing strategic competition.
For defense analysts and policymakers in Washington, the AEI report’s most important finding may be the one about institutional constraints. The PLA’s cultural resistance to autonomous decision-making, its preference for centralized command and control over the kind of machine-speed autonomy that AI-enabled swarming requires, is a genuine limitation on its military AI capabilities. But it is also a limitation that is likely to erode over time, as a new generation of PLA officers who have grown up with AI systems rises through the ranks and as the operational evidence from future conflicts continues to demonstrate the decisive advantages of autonomous systems. The window in which this constraint provides a meaningful advantage to US and allied forces may be shorter than current planning assumptions suggest.