AI-generated content: authorship and inventorship in the age of artificial intelligence

Predicting the future of technology is notoriously difficult. Indeed, predicting how law and regulation should be shaped to meet the needs of future technological developments is a task that might often lead to hilarious predictions.1 The difficulty in predicting technological development is certainly reflected in the current debate about the future of artificial intelligence (AI). Within this framework currently two extremes can be identified: those who view AI as a path towards ‘superintelligence’ that transcends humanity, and those who think AI is merely a glorified version of data analysis and statistical inference. In any case, it seems realistic to foresee that in the near future there will be an increase in machines that are able to perform more tasks in more efficient and autonomous ways than we currently can envision. These tasks include the production of artistic, technological, and scientific innovations that might potentially be protectable via intellectual property (IP) laws. Because of the economic value of these innovations there may be an interest in ‘controlling’ such intellectual creations via intellectual property right (IPRs). In this context, a key question relates to how to interpret the concepts of ‘authorship’ (copyright) and ‘inventorship’ (patents) of creations and inventions generated by AI systems. IPRs aim at protecting the fruits of the human mind. IPRs are a set of limited exclusive rights allocated to ‘persons’, either natural or legal. As such, traditionally, both copyright and patent laws in Europe have relied upon the concept of author or inventor as a natural person. Indeed, such an idea is reflected not only in the legal definitions of author and inventor, but also in concepts like ‘work’ and ‘invention’, as well as in the requirements necessary to acquire protection. By possibly allowing production of innovations in an ‘autonomous’ way, AI naturally challenges these traditional well-established legal notions. Can an AI-generated work or invention attract IPR protection under current rules? Is there a need to shape our understanding and interpretation of authors and inventors as natural persons in view of rapidly expanding AI? What would be the benefits and what the consequences of such shift? The chapter focuses on the concepts of authors and inventors for AI-generated innovations as prerequisites for copyright and patent protection. The issue is important not only because authors and inventors have ownership rights in their copyright or patent, but also (and more fundamentally) because if there is no legitimate author or inventor, the IPR might either not be granted in the first place or held invalid or unenforceable in court. The chapter begins by explaining some fundamental technological concepts of AI and by providing examples of AI-generated outputs that can potentially attract IPR protection. The second part explores copyright and patent