Artificial Intelligence-Generated Music and Copyright: Regulatory Challenges and Legal Perspectives

Abstract

This study investigated how current copyright legislation addresses music generated by artificial intelligence (AI) through a systematic review conducted in the Scopus and Web of Science databases, covering the period from 2015 to 2025 and based on the PRISMA method. A total of 83 articles were identified, of which 13 met the inclusion criteria and composed the final sample. The results indicate that the copyright legal framework remains grounded in human centrality, requiring the participation of a natural person in the creative process as a condition for protection. Although part of the literature acknowledges more flexible interpretations, especially in contexts of greater system autonomy, gaps persist regarding the definition of originality, intentionality, and the legal role of prompts and training data. In the regulatory domain, three main approaches stand out: the proposal of a sui generis legal regime, the adoption of incremental adjustments to existing norms, and the implementation of compensatory mechanisms aimed at redistributing value. In light of these findings, it is concluded that the current copyright model does not adequately address AI-generated musical works, requiring the development of normative solutions that reconcile human centrality with emerging technological dynamics, under the risk of legal uncertainty and imbalance in the creative ecosystem.