Analysis of Data Access Restrictions by AI Providers from a Competition Law Perspective: Considering the Objectives of Copyright Law

Abstract

This study analyzes the data access restrictions imposed by artificial intelligence (hereinafter “AI”) providers from both antitrust and copyright law perspectives, focusing on the intersection of the two regimes. Because training data are essential for developing AI, companies may have incentives to limit competitors’ access to data. Drawing on foreign regulations, precedents, and theories of harm, this study examines how such restrictions can be assessed under Korea’s Monopoly Regulation and Fair Trade Act (hereinafter “MRFTA”), specifically regarding the abuse of market dominance and unfair trade practices. When data holders assert exclusive rights under copyright law, such rights may justify data access restrictions and potentially exempt the conduct from antitrust scrutiny under the MRFTA. Accordingly, this paper further explores the IP rights and implications of data training and access restrictions, introducing guidelines from two jurisdictions that clarify the boundary between justifiable exercise and the abuse of IP rights. The analysis emphasizes the consideration of market dynamics and consumer welfare when evaluating the exercise of IP rights and calls for a balanced legal framework that protects both innovative incentives and competitive market structures. This study contributes to ongoing legal and policy debates on reconciling the purposes of intellectual property law and competition law in AI and data governance.