인공지능 주권 논의와 데이터 법⋅정책에 관한 비판적 고찰

Abstract

The third golden age of artificial intelligence has begun. Unveiled in late November 2022, Chat GPT, an ambitious work of open AI, is a conversational AI model designed to interact with humans. Generative AI is a technology based on language models that learns from good data to answer a variety of questions and generate new answers. Today, global big tech companies are the largest holders of data that is key to AI learning, and much of it is accumulated from existing platforms and is copyrighted work. As a result, the collection of data against the will of the right holders is problematic, and disputes are intensifying. Data is the crude oil of the informationization industry, and its economic value has become even more important in the AI environment. Conscious of the monopolization of data by US companies, countries such as the European Union, France, and China have responded with their own data protection policies, and the situation is similar in Korea. So-called “data sovereignty,” “digital sovereignty,” “information sovereignty,” and the recently emerged “AI sovereignty” can be misused to protect and control their own data and the data of their citizens behind this technological hegemony. The inherent value of data is to create economic value through distribution, and the internet is a space where data can be freely distributed. Therefore, control-oriented data protection policies are in conflict with the nature of data. In Korea, while companies and government projects that are late in developing generative AI are looking for ways to share data, they are faced with the dilemma of adhering to policies to protect K-content and strengthen copyright in response to global big tech, and are calling for sovereign AI. This confusion stems from the lack of a legal foundation for technology sovereignty and a national consensus on data policy. Based on the legal issues that have emerged in the conflicts between individuals, companies, and countries over data in the era of artificial intelligence, This paper analyzes the legal issues that have emerged in the conflicts between individuals, companies, and states over data in the era of artificial intelligence, and critically approaches the recent claims based on 'technological sovereignty' from a legal perspective. As a result, we found that 'technological sovereignty' is a rhetoric rather than a legal concept, and that the sovereignty of the most important information subject, the user, is missing from the discussion. From a legal perspective, information actors in the technological environment are individuals, companies, and countries, and users are active participants who drive the quality of technology and services. Involving users as autonomous subjects in the discussion of technological sovereignty, with common goals and values shared by the state, companies, and citizens, will contribute to national industrial competitiveness in the dilemma surrounding data.