The Global AI Full-Stack Race and Korea’s Response: Beyond AX Adoption to Supply Capability

South Korea is actively spreading artificial intelligence (AI) transformation (AX) among its manufacturers, but the core supply base that drives it depends on foreign technology. Global competition has moved beyond individual technologies, the separate layers of the stack, to a contest of full-stack vertical integration for control of every layer of the stack, from energy to applications, and platform lock-in is entrenching the dominance of a handful of fi rms. In this setting, spreading AX without regard to the supply side could end in dependence on foreign platforms: a replay, on the AX stage, of Korea’s materials, parts, and equipment crisis. This paper looks for a way through in what distinguishes the AI ecosystem from the AX ecosystem. The upper layers of the stack, which produce general-purpose intelligence, are a contest of capital and scale in which catching up is not realistic. By contrast, industry-specifi c AI (vertical AI), which implements that intelligence in industrial settings, and the capability to supply it are where Korea, with its accumulated manufacturing experience, can realistically compete. The paper analyzes four structural limitations of Big Tech’s focus on general-purpose AI and reviews developments in the major economies. It then proposes a response that moves beyond AX adoption toward stronger supply capability: pursuing demand and supply together and staking an early claim to operational standards.