현행 소비자보호법제에서 인공지능 기술 적용 가능성과 한계

Current consumer protection legislation, premised on the rational and average consumer, has encountered the limitations of bounded rationality and cognitive biases presented by behavioral economics. In particular, traditional regulatory frameworks alone struggle to effectively counter new threats like dark patterns that deliberately exploit cognitive biases. This presentation examined the applicability of AI as a normative and technological alternative to bridge these legal gaps. AI technology holds significant potential to impact consumer law systems broadly as a regulatory technology and legal tech. Specifically, AI can be utilized in the area of contract law for automated screening of unfair clauses and detecting deceptive UI/UX and linguistic deception in dark patterns. Furthermore, in the realm of the Labeling and Advertising Act, it enables verification of false or exaggerated advertising, identification of undisclosed compensation for influencers, and proactive detection of product defects from unstructured data linked with systems like CISS. Its utility extends broadly to automating dispute resolution procedures, assisting negotiations, and supporting evidencebased policy formulation. However, these technological possibilities may encounter legal and institutional limitations. The issue of AI data bias can lead to regulatory outcomes disadvantageous to specific consumer groups and entrench bias through feedback loops. Furthermore, AI’s black-box nature, with its inability to explain decision-making grounds, risks violating the duty to provide reasons under administrative procedure laws, infringing on the defense rights of regulated parties, and rendering judicial oversight ineffective. Moreover, it may trigger technological competition, such as businesses intelligently circumventing regulations or attempting personalized deception. Other threats to AI’s effectiveness include conflicts with personal information protection laws, automated biases within regulatory agencies, and the technological gap between large platforms and regulators. While AI is clearly a meaningful tool for enhancing consumer protection effectiveness, careful consideration is needed to harmonize the conflicting values of regulatory fairness, procedural legitimacy, and data protection versus utilization.