인공지능(AI) 윤리원칙을 중심으로 한 규제 내용 분석·비교

Abstract

As artificial intelligence (AI) technology and associated products and services become more common, ethical issues pertaining to opacity, bias, and privacy continue to be raised, and concern is growing. Therefore, various discussions are underway regarding the ethical implementation and regulation of AI so that it can be reliably regulated on an international scale. Recently, the focus of the discussion has changed from building norms regarding AI's usage to the drafting and passage of legislation that regulates it. Therefore, this paper examines recent trends in major countries related to the regulation of AI, and analyzes and compares these countries' approaches to AI legislation and regulation, focusing on AI ethical principles. This discussion describes the direction in which the international community is defining and regulating the ethical use of AI, and it suggests regulations that are consistent with global norms and but also appropriate for the domestic situation in Korea. Analysis reveals that international discussions are suggesting practical approaches to the identification and mitigation of AI's ethical risks. Trends in legislation and regulation suggest that the international community is prioritizing the maintenance of ethical principles within the context of AI implementation. While the United States, EU, and China have regulated AI in a manner consistent with the principle that ethical and legal standards should be centered on people, the three bills currently presented to Korea's National Assembly do not uphold that standard. The legislation has not sufficiently respected common standards of fairness, accountability, privacy protection, data management, and technical safety. In order to remedy this deficiency, this paper suggests improvements to a general direction for future domestic AI regulations along with specific improvements to legal provisions.