Inventorship in the age of AI: examining the USPTO Guidance on AI-assisted inventions

Abstract The rapid rise of artificial intelligence (AI) in innovation has introduced challenging questions at the intersection of patent law and AI, including inventorship. In response, the United States Patent and Trademark Office (USPTO) issued its 2024 ‘Inventorship Guidance for AI-Assisted Inventions’, emphasizing that inventorship is human-centric. An AI system cannot be named an inventor, although it may assist in a patentable inventive process if a natural person significantly contributes to the invention’s conception. This position is grounded in statutory language, judicial precedent and the view that patents are intended to recognize and reward human ingenuity. This article examines the USPTO’s Inventorship Guidance, analysing its guiding principles and examples for AI-assisted innovation. Key elements include the ‘significant human contribution’ standard, which requires that a person’s role goes beyond simply operating or interpreting an AI system’s output. We also discuss the role of the Pannu factors—longstanding criteria for joint inventorship, re-purposed in the Guidance to assess whether a human meets the threshold of an inventor when AI has assisted the development of the invention. For practitioners, the guidance provides greater clarity on inventorship requirements in AI-assisted inventions, reinforcing the requirement for significant human involvement in their conception. Important practical lessons emerge. In addition, a variety of deeper questions remain open for further research.