AI Inventorship and Patent Law

In our rapidly advancing technological era, we are witnessing remarkable strides in various fields of innovation. Notably, the realms of science and technology are being transformed by cutting-edge technologies such as Artificial Intelligence (AI), Blockchain, and 3D printing. These advancements are setting new benchmarks in reducing human involvement in complex processes.


AI, in particular, stands out as a groundbreaking technology. It involves machines, primarily computers, mimicking the human cognitive process. The evolution of AI has enabled these machines to absorb knowledge from experiences, adapt to new scenarios, and execute tasks that were traditionally human-centric. As a result, AI systems, much like their human creators, are now capable of creating inventions, thus positioning them as potential inventors. However, this raises an important question: how do AI inventions fit within the framework of patent law, especially with the United States Patent and Trademark Office (USPTO)?


This issue first caught the attention of the USPTO in 2019, when they sought public feedback on patenting AI-generated inventions. The main concern was whether the existing patent laws were adequate for AI-driven creations. Most of the feedback highlighted the need for new policies to
acknowledge and safeguard inventions produced by AI.


A significant case that put these questions into perspective involved Stephan Thaler and his AI
system, DABUS (Device for the Autonomous Bootstrapping of Unified Sentience). In 2018, DABUS designed a novel beverage container and a warning light for autonomous vehicles. Thaler, seeking patents for these inventions in various jurisdictions including the US, UK, and Europe, named DABUS as the sole inventor. These applications were rejected on the grounds that inventors must be natural persons.


Thaler challenged this decision in the Eastern District of Virginia. The court, referencing 35 U.S.C. 100(f), declared that an “inventor” must be a “natural person,” effectively ruling out AI systems as inventors. Thaler appealed, but in August 2022, the Federal Circuit upheld the original ruling, asserting that the term “individual” in the patent law context referred explicitly to human beings.

The Thaler v. Vidal case clarified that AI cannot be legally recognized as an inventor in the context of patent law. This decision was met with disappointment in the tech community. The argument is that if the patent law, as outlined in Article 1, Section 8, Clause 8 of the US Constitution, is intended to advance science and the useful arts, then excluding AI as inventors could hinder businesses that rely on AI for innovation. The primary goal for patent law is to incentivize inventors by granting them exclusive rights for a limited period in return for disclosing their invention. Without AI recognition, companies may feel discouraged from pursuing AI-driven self-innovation.


The debate extends to the notion of “non-obviousness,” a core patent criterion. While AI inventions may be novel, their obviousness is assessed differently compared to human-generated ideas. What might be non-obvious to a human (POSITA) might not be the same for an AI system, which learns and operates based on data sets and machine learning algorithms.


To address these challenges, it is proposed that patent laws be reformed to recognize AI inventorship. This could involve categorizing inventions into those created by humans and those by AI systems. This way, when filing patent applications, the contribution of both human inventors and significant AI systems can be accurately acknowledged. Such an approach would prevent the misrepresentation of inventorship and encourage companies to continue innovating with AI, knowing they can secure patent protection. The public, in turn, benefits as these inventions, once patented, eventually enter the public domain.


For the advancement of science and technology, it is critical that Congress and the USPTO consider and adopt amendments to existing patent laws. These changes should accommodate AI inventorship, ensuring that patent attorneys, inventors, and AI enthusiasts alike can navigate this new frontier with clarity and confidence. D’Ambrosio & Menon is a trusted law firm, specializing in patent law and intellectual property. Contact us today for expert legal counsel.