AI: Artificial Intelligence or Artificial Inventorship?

With the recent advent of generative AI, many have approached its capabilities with mixed feelings; feelings ranging from curiosity to wariness. In fact, many have compared the advent of gen AI to the advent of the steam engine, which eventually propagated the industrial revolution of the 18th century. Others have compared the advent of gen AI to the internet boom of the 90s.  Regardless, much like many other technological advances, it appears that the younger generations, particularly Millennials, Gen Y, and Gen Z generations, are not only more adept with respect to its use, but also embrace it with a positive outlook. 

Gen AI is now widely known for its ability to generate creative works of music, text, code, image, and video. However, gen AI is also known for its propensity to "hallucinate” causing it to generate false historical events, false citations, or even disturbing and creepy images. But what happens when programmers and software developers are able to harness the power of gen AI to create software codes, programs, games, or apps? What about research scientists who have mounds of data to sift through to find the proverbial needle in a haystack? Or engineers who are taxed with formulating complex structure designs with mathematical precision? And how is the law, with respect to gen AI, involved at the intersection of software, science, and engineering advancements? One answer is intellectual property law. 

IP law, which encompasses trademark, copyright, trade secrets, and patent law, like other forms of property law, is most concerned with ownership rights and the right to exclude others with respect to a particular invention or design. Patent law, in particular, is so concerned with ownership rights, that the name of the applicant, inventor, and/or assignee is published in bold on the face of each patent and patent application. Yet until a few years ago, there was little confusion as to who would be listed as an inventor since guidance was clearly laid out in the Manual of Patent Examining Procedure (MPEP), section 2109. To make things more complicated, queue in gen AI. Or more appropriately, “prompt” in gen AI. 

Guidance under the MPEP, section 2109, subsection II, defines inventorship as an individual who contributes to the conception of the invention. While subsection IV describes joint inventorship based on at least some contribution to the claimed invention by each of the named joint inventors. But prior to revisions in 2023, there was no guidance limiting joint or named inventorship to human beings, and it appeared that one could freely claim AI as a joint inventor.

So what would happen if AI wrote a code that contributed to the conception of the invention – would AI be listed as a joint inventor? Is AI considered an individual under the broadest interpretation of the statutory language? The Federal Circuit Court addressed these issues in its 2022 Thaler v. Vidal decision. 

In Thaler v. Vidal, the sole issue on appeal was whether an AI system could be considered a joint inventor on a patent. The Court’s conclusion that the statutory language for inventorship is “limited only to natural persons” ultimately led to a revision of the MPEP, section 2109, to add subsection VII, which limits inventorship to a natural person, eliminating a lot of confusion from the patent landscape.

However, the United States Patent and Trademark Office (USPTO) has recently addressed a more important question: can AI-generated inventions be patented? In particular, can a human inventor patent an AI-generated invention if it was prompted by minimal effort from said inventor? 

In response to the nebulous landscape of AI in patent law, the Federal Register published a notice on February 13th, 2024 entitled “Inventorship Guidance for AI-Assisted Inventions.” The notice provides preliminary guidance for examiners and patent practitioners alike in navigating AI in patent law. One major takeaway that helps determine whether AI inventions can be patented is that the USPTO applies the Pannu factors of contribution to discern inventorship and contribution from AI. In the Pannu factors, or the “significant contribution test,” inventorship rights essentially hinge on a natural person’s contributions to the claimed invention.

As such, when a natural person uses AI for an invention, the named inventor must be a natural person and must have also significantly contributed to the claimed invention. Ultimately, the USPTO allows inventors to freely use AI tools to aid in the creation of an invention so long as contribution efforts have been appropriately identified and satisfied according to the guidance provided. 

Thus, an invention, where burdensome efforts can be offset by the use AI tools, AI can be particularly helpful. While there is no brightline rule for identifying specific contributions from AI, the USPTO guidance identifies a list of factors that may help discern any significant contribution efforts. 

Ultimately, the use of generative AI tools has not been prohibited by the USPTO in patenting an invention so long as a human being significantly contributed to the creation of the invention. Afterall, the patenting system has been carefully designed to promote human ingenuity, not artificial creativity. Although AI technology is likely to continue advancing, much like the steam engine or the internet, any technological advancement or revolution still relies on significant human contribution. At least for now.

Israel Suarez is a second-year law student at Wake Forest University of Law. A proud Double-Deac, he holds a Ph.D. in Organic Chemistry from Wake Forest University. He also holds a B.S. in Biology from Catawba College in Salisbury, North Carolina. Israel plans on pursuing a career in patent prosecution in the chemical and life sciences field following law school.

Reach Israel here:

LinkedIn: https://www.linkedin.com/in/sisraelsuarez

Email: [email protected]