Do AI-generated outputs constitute a copyright infringement? : a Maltese legal perspective

Abstract

This dissertation examines the implications of generative artificial intelligence for copyright law, with particular focus on the Maltese legal framework within the broader context of European Union law. It addresses the central research question of whether AI-generated content can constitute copyright infringement and, if so, who may be held liable under existing legal regimes. The analysis is grounded in a doctrinal (nonempirical) methodology, relying on a detailed examination of legislation, case law, and academic commentary. The study evaluates the applicability of key EU instruments, notably Directive 2001/29/EC (InfoSoc Directive) and Directive (EU) 2019/790 (Directive on Copyright in the Digital Single Market), as well as relevant jurisprudence of the Court of Justice of the European Union. Particular attention is given to concepts such as authorship, originality, communication to the public, and the scope of text and data mining exceptions, assessing their relevance in the context of generative AI systems. The findings indicate that, while existing copyright frameworks provide partial guidance, they are not fully equipped to address the novel challenges posed by AI technologies. Uncertainty persists regarding the legal qualification of AI-generated outputs and the attribution of liability among developers, deployers, and users of such systems. The dissertation concludes that greater legal clarity is required, potentially through legislative reform or judicial interpretation, to ensure that copyright law remains effective and balanced in the face of rapid technological advancement.