Abstract

There are numerous problems with permitting AI systems to be inventors for patent purposes. These problems include creating issues with the analogous art requirement, failing to meet the enablement standard, recalibrating who the Person Having Ordinary Skill in the Art is, generating constitutional concerns about incentivizing AI, producing similar incentives to have AIs treated as the authors of copyrighted works, and setting the stage for other non-human entities to have intellectual property rights. These problems need to be answered before AI is allowed to be an inventor on patents. This Article addresses these concerns, then proposes a solution: having the discoverer of the AI’s invention be the “inventor.”

It should be noted that this paper makes a key assumption about AIs. At least a few scholars have thought of AI-inventorship as a spectrum. At one end of the spectrum is the sole inventor working without the benefit of AI. At the other end of the spectrum is AI-solely inventing with no input or direction from humans. In the middle of this vast spectrum is an increasing amount of AI-contribution. This Note examines the end of the spectrum where AI is the primary inventor. The invention process could involve some input or direction from humans, but AI performs the conception step. This is important because conception is often considered the “touchstone of inventorship.” Conception “is the formation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention, as it is hereafter to be applied in practice.” It should be noted that this Note does not address AI-generated claim sets.

How to Cite:

Pressley Nietering, Why Artificial Intelligence Shouldn't Be a Patent Inventor, 5 Ariz. L. J. Emerging Tech., no. 5, 2022, https://doi.org/10.2458/azlawjet.5510