한국의 AI 기술경쟁력 강화를 위한 데이터 주권 확보 방안

Digital sovereignty refers to a condition in which a state or community exercises autonomous and substantive authority over the development, use, control, and rule-making of digital technologies. It is a holistic concept that goes beyond merely possessing technology to encompass data sovereignty, AI model sovereignty, and normative sovereignty over policies and ethics. Among these, securing AI technological competitiveness is regarded as a core prerequisite for digital sovereignty. If a country cannot design and operate AI models on its own, sovereignty over data and policy will inevitably ring hollow. Indeed, in the early digital era of the 1980s, Microsoft’s operating-system monopoly—and, subsequently, Google’s dominance of search algorithms—demonstrated how the loss of technological sovereignty can extend its effects to user experience and even policy domains. Today, with the advent of generative AI and large language models (LLMs), structures of technological dependence are being reconfigured in more sophisticated and pervasive ways. As the AI services of a small number of firms—such as GPT, Claude, and Gemini—consolidate platform power worldwide, South Korea likewise faces the risk of “digital colonization” by global AI Big Tech. Against this backdrop, this paper analyzes Korea’s AI technology–dependence structure and argues that the most urgent prerequisite for any strategic approach to securing data sovereignty is to break away from path dependence on Big-Tech AI technologies. Building on this premise, it proposes core strategies for Korea to secure AI technological competitiveness—namely, achieving the self-reliance of homegrown AI models through data sovereignty and establishing an independent build-out of AI infrastructure.