작곡 AI의 현재와 쟁점에 따른 국내 음악저작권 논의방향 연구

Abstract

AI composer stands out as one of the most vigorously applied domains of AI technology within the music sector. This study discusses the current status of AI composer as perceived by various stakeholders in Korea through recent issues in the music industry and relevant literature, while also examining the copyright-related issues and directions of discourse. The aim is to encourage policymakers to adopt a balanced perspective on technological advancement and copyright protection, and to facilitate understanding among music industry stakeholders and composers regarding the direction of copyright discussions. To emphasize the contentious issues arising in the process of music content production, this study categorizes them into pre-creation stages, creation stages, and other areas. First, in the pre-creation stages involving data learning, copyright infringement issues arise from the replication and transmission of original data. Thus, we propose the necessity for transparent disclosure of information regarding training data, as well as discussions on the extent and method of compensation for the data copyright holders, and the need for policy support to facilitate such agreements. Second, in the post-creation stages, copyright issues revolve around the recognition of AI composer creators, ownership attribution, and protection methods. Concerning the recognition of creators, as instances of recognizing AI as creators increase globally, this study suggests preparations for potential impacts on international copyright laws. Furthermore, the distribution criteria based on AI’s role and contribution must be established, with an emphasis on the necessity for internal consensus within the music industry. Lastly, the sharing and commercial use of music composed by AI and the determination of similarity with existing works are also discussed. Specifically, ethical guidelines regarding plagiarism and the need for systems enabling objective judgment are proposed.