<![CDATA[The hegemonic rivalry between the United States and China has extended into the domain of Artificial Intelligence (AI), catalyzing a new autonomous arms race. This article examines how military AI activates the security dilemma and undermines global stability. Grounded in Kenneth Waltz’s Neorealism and Robert Jervis’s offense-defense differentiation, this qualitative study utilizes process tracing and critical documentary analysis. It reveals that the intangible and dual-use nature of AI erodes the boundary between defensive and offensive postures. Consequently, integrating AI into warfighting doctrines—such as the U.S. JADC2 and China's Intelligentized Warfare—fuels a spiral of uncertainty, severely compresses human command deliberation, and heightens the risk of algorithmic flash wars. At the macro level, battlefield transparency compromises nuclear deterrence stability, while multilateral diplomacy in Geneva remains deadlocked. This study concludes that the pursuit of algorithmic military superiority engenders profound strategic fragility, underscoring the urgent need for institutionalized confidence-building measures and a binding taboo against AI in nuclear command.]]>
Copyrights © 2025
