Going Deep: EU Copyright, Generative AI and the Competition Rationale Underlying Originality

In the current debate on the copyrightability of AI production, several arguments were brought in relation to the foundations and rationales of copyright law. Against this theoretical background, a clear line was drawn between AI-generated and AI-assisted production. Whereas it seems now generally admitted that copyright will not vest with the former because of its complete lack of authorship, the latter seems eligible to such protection as there is here room for human intervention. Though human authorship is a bedrock requirement of copyright, it does not however suffice to conclude in general that AI-assisted productions would fall within its realm. Other rationales justify the grant or denying of this protection. Amongst those, the one underlying the US merger doctrine, and that can be found in the CJEU case law (Copyright, Design, Trademark), shows that copyright protection is probably not fit for AI-assisted productions per se. This idea, that connects the competition foundations of copyright and the basic technical features of Generative AI, was apparently left unexplored. Enshrined in IP theory, it exemplifies that the New Digital Rule of Law is not simply a New Rule of Law in the Digital.