Ars Ex Machina : Originality and Authorship in Computer-Generated Works in European Copyright Law

Abstract

Artificial intelligence and machine learning systems are today capable of autonomously and independently generating works that would be copyright protected had they been created by a human. The rise of computers capable of creating works of art are challenging the notion of creativity being exclusive to humans and creates a copyright conundrum, since copyright law only recognizes human authors. The purpose of this thesis is to examine whether computer-generated works can be copyright protected according to European Union copyright law, and if the answer is yes, who the author of a computer-generated work is? The methodology used to answer these questions is legal dogmatic, and the research question will be approached through originality and authorship according to EU copyright law. The conclusion is that the most suitable author of autonomously computer-generated works, following the EU originality criterion, in most cases is the computer itself, but since non-human authors are not recognized these works seem to fall outside the scope of copyright protection. The current EU copyright system is not equipped to handle autonomously computer-generated works, and some proposed or implemented solutions to this conundrum, e.g. the UK provision on computer-generated works, are not necessarily compatible with EU copyright law.