Responsibility for AI Inventions

Responsibility for the activities and products generated by artificial intelligence (AI) is a topic currently debated by scholars. However, less attention has been given to the case of inventions created by AI. Who should be held responsible for these inventions? How can we address the negative consequences arising from AI-generated inventions? In this article, we aim to highlight some unique aspects of responsibility related to this specific type of invention. We examine the possibility of addressing responsibility for AI-generated inventions through the concept of "legal personhood," which we conclude introduces a new set of challenges arguably even more complex to resolve. Therefore, we propose several governance tools and outline two practical approaches that have been attempted. This paper offers a tentative discussion on a novel topic within social AI research.