‘The rapid proliferation of generative artificial intelligence (GenAI) technologies has intensified debates in copyright law regarding the scope of exclusive rights, the role of exceptions and limitations and the allocation of liability among the multiple actors operating within complex technological value chains. This article maps and critically examines the principal copyright liability-related issues currently emerging under EU law as a result of the interaction between the Directive on Copyright in the Digital Single Market, the Artificial Intelligence Act and the broader EU copyright acquis. The analysis is structured around three core phases of the GenAI value chain: (i) the training phase, (ii) the storage and distribution of trained models and (iii) the generation and dissemination of outputs. In doing so, it analyses the relevance and limits of the text and data mining exceptions, the legal significance of model “memorization” and the qualification of potentially infringing outputs. Beyond this structural analysis, the article identifies a set of contentious issues that are increasingly surfacing in litigation and regulatory debates. These include the legal implications of retrieval-augmented generation, the potential infringement of moral rights, the circumvention of technological protection measures and the evolving framework governing remedies for copyright infringement.’
-
Subscribe
Subscribed
Have a WordPress.com account? Log in now.
