인공지능에 의한 저작물 사용 관련 저작권 침해 책임 면책과 저작권자에 대한 보상에 관한 연구

The emergence of generative AI has raised new issues for both AI developers and copyright holders. These include copyright infringement issues that may arise when copyrighted works are used to train AI (including generative AI), the absence of a realistically feasible licensing system for AI developers, and the lack of transparency regarding what copyrighted works are used in AI training and to what extent they are used. For these reasons, this article seeks to explore several legislative, judicial, and policy solutions that can overcome these issues, with the goal of balancing the demands of both the AI ​​industry and the copyright industry in Korea, as copyrighted works are used in generative AI training. As specific measures, regarding the issue of exemption from liability for copyright infringement related to the use of copyrighted works by AI and compensation for copyright holders, it examines the need to distinguish between “creative AI” and “non-creative AI,” recognize an exception to copyright property rights for “non-creative AI” and grant statutory license for “creative AI,” and proposes imposing obligations of transparency regarding the use of copyrighted works on AI developers. The series of solutions presented in this article aim to establish a copyright system that encourages rather than suppresses the development of AI technology. In addition, it is expected to be used as a theoretical basis for legislative data to achieve a balance between the development of the AI ​​industry and the copyright industry in the future, by aiming to provide fair compensation to copyright holders for the use of human-created works when such works are used to train AI to produce outputs (i.e., content) that compete with or replace human-created works, in order to protect the legitimate interests of copyright holders.