Autonomy of AI in Patents: Reconciling Commercial Incentive with Traditional Inventorship

Abstract

This paper examines the general commercial implications of denying inventorship to an artificial intelligence (AI)
programme which functions independent of its human creator and generates patentable products or processes. Traditional
rules of inventorship pre-suppose that an inventor must be a natural person capable of exercising rights and being bound by
duties and liabilities in relation to the patented invention. Thus, AI becomes automatically disqualified and cannot be named
as an inventor in a patent application for this small technicality. Recently, patent applications filed by Dr. Stephen Thaler for
inventions by his AI DABUS, have been rejected due to this hurdle. The article highlights the need for clarity in the area of
patent inventorship for AI. Even when objections are raised based on traditional patent law principles, one must not forget
that historically, commercial incentive has been the core objective of every patent legal system. Thus, encouraging
applicants to step forward with new and useful innovations has always remained at the forefront of patent laws worldwide.
The article elaborates on this conflict and suggests a suitable solution that appears satisfactory to all stakeholders