The adoption and development of AI technology in the defense sector is now an unstoppable trend, driven by the ‘Global AI arms race’ between military superpowers like the United States and China, as well as the projected rapid growth of the milita...
The adoption and development of AI technology in the defense sector is now an unstoppable trend, driven by the ‘Global AI arms race’ between military superpowers like the United States and China, as well as the projected rapid growth of the military AI market. These trends will bring significant changes to the traditional defense industry ecosystem and defense R&D. Therefore, this study investigated the structure of defense AI research collaboration, the convergence and diffusion of AI technologies with military technologies, and the potential military applications of AI sub-technologies. To this end, we extracted and analyzed defense-related AI patent information filed in South Korea, the United States, and China between 2015 and 2023. The results show that defense AI patent applications in South Korea are increasing, while those in China have grown rapidly, surpassing the United States in the number of applications since 2022. Analysis of research collaboration networks based on joint patent applications indicates that South Korea, the United States, and China all maintain fragmented and closed research collaboration structures similar to traditional defense R&D. In South Korea, defense AI research collaboration is expanding, but it remains concentrated in partnerships between defense research institutes and universities, and between defense companies and universities. In contrast, the United States has predominantly enterprise-centered collaboration structures, whereas China’s collaboration patterns are predominantly university-centered. An analysis of technology convergence revealed that both South Korea and China have continuously observed the emergence of new AI and military technologies, with the scope and intensity of convergence increasing over time. In South Korea’s defense sector, the convergence of AI sub-technologies with military technologies is expanding, thereby increasing the potential for military applications of AI. AI sub-technologies with high military applicability largely overlapped with promising AI technologies previously predicted by experts, although some of these technologies were found to require technology transfer from the civilian sector. This study contributes to the advanced field of defense AI by proposing various patent-based methods to analyze research collaboration structures, technology convergence, and technology forecasting. These methods can also be applied to other areas of advanced defense technology, thereby supporting more effective defense R&D.