Comparative analysis of patent law systems in the European Union and the United States of America in the context of patenting of Artificial intelligence and its outcomes.

Abstract

AI has been used as a tool to speed up innovation for many years and whereas previously was entirely relying on its operator to at least insert initial data, the modern development of arising technologies allows AI being fully capable of generating inventions by themselves. It has been argued that AI-generated inventions have been submitted all around the world for at least a decade now, which raises some questions. What is the future of AI-generated inventions? Can an invention, fully concluded by an AI, receive proper recognition based on the legislation of the EU and US? This paper argues that modern patent system development precludes from the proper acknowledgement of AI as an inventor, as patent law predominantly centres around the natural person. This creates many complications in regulation and potentially a grey zone in legislation that allows abuse of the patent system. Some cases are showing how patent law has been interpreted differently through years between EU and USA despite in essence deriving from the same source; and at the same time how much regardless all the efforts, it is far from being unified. Through the historical and analytical approach, it is argued, that the modern patent system is not suited for the newly arisen technologies and needs to be amended in order to regulate them. The problem of regulation AI-generated inventions should not be approached from a patent law perspective only, as it is more complex and must include also a question of liability.