LIABILITY FOR AI-GENERATED ART: AUTHOR OR ALGORITHM?.

Abstract

The article investigates pressing legal issues arising from the use of artificial intelligence in generating artistic works. Particular attention is paid to the question of legal liability for the results produced by generative systems, namely whether such liability should rest with the algorithm itself or with identifiable human actors such as developers, providers, or users. The author analyses legislative approaches in both the European Union and Ukrainian legal systems. Significant attention is devoted to scholarly concepts of the legal personality of artificial intelligence, viewed through the lens of the electronic–person theory, to case law on copyright in AI–generated content, and to the provisions of the EU Artificial Intelligence Act. It is demonstrated that, under current conditions, an AI system remains an object of civil commerce rather than a subject of law; therefore, legal liability in the artistic sphere for AI–generated content is borne by natural persons in proportion to the degree of control they exercise over the system. The characteristics of AI examined in the article align with those of a thing or technology under civil–law doctrine, so the algorithm itself cannot be a subject of civil liability in the arts. To mitigate the risk of infringements, particularly in copyright, the article emphasises the advisability of implementing technological protection measures. It concludes that the normative establishment of differentiated liability corresponding to each participant’s role in creating and using AI content remains contested and requires resolution, a task complicated by the rapid evolution of AI–related relations and the absence of a unified global approach.