If the Inventor is No Longer Human, Then What?: How the USPTO is Addressing AI Assisted Inventions

Ivette Rodriguez Nieves, Contributing Member 2024-2025
Intellectual Property and Computer Law Journal

I. Introduction

Artificial intelligence (AI) has become hard to escape in today’s environment and has led to many companies to implement and capitalize on the movement. Prominent industries such as technology, pharmaceuticals, and healthcare are utilizing AI technology to help accelerate innovation.[1] Hastened innovation provides a crucial benefit to inventors who wish to patent their innovationsbefore competitors.[2] Generally, Art. I Sec. 8 Cl. 8 of the United States Constitution grants inventors with temporary market exclusivity in exchange for the disclosure of specific details of the invention.[3] However, it has become more ambiguous as to whether a person who uses generative AI in an invention, can be granted protection.[4] For example, the Federal Circuit has determined that generative AI is not human, therefore it cannot be an author.[5]

Part II provides significant background on key factors the legal system considers to be essential to this new landscape for granting patents to inventors. Part III discusses the effect the guidelines have on patent applications and potential issues that could arise. Part VI concludes by contemplating the USPTO’s involvement in addressing inventorship issues.

II. Background

Under Art. I Sec. 8 Cl. 8, Congress is granted with the power to promote the progress of science and useful arts by providing authors and inventors the exclusive right to market their invention without competition their respective writings and discoveries for a period of time.[6] However, the word ‘inventor’ is difficult to interpret with the emerging use of AI in the discovery process.[7] Recently, there has been a rise in inquiries of: who qualifies as an inventor and what limitations exist to be an inventor.[8]

When a person inputs information into a generative AI system that returns vast amounts of information, it becomes difficult to draw the line of who is the true creator of the invention.[9] This blurred line has caused patent examiners to struggle with granting exclusive rights.[10]

With the increased integration of generative AI in research, the Biden Presidential Administration issued an Executive Order in October 2023.[11] The Order urges the USPTO to set guidelines for inventors and applicants on what may be patentable and what may be unpatentable.[12] The USPTO, then indicated that using AI in an invention is not unpatentable, but listing it as an inventor is unpatentable.[13] Additionally, a factor in a patentable invention that will be addressed further in the article, must be that a human being significantly contributed in the making of the invention.[14]

USPTO and The Guideline

The USPTO is the federal agency for registering trademarks and issuing patents.[15] It fulfills Art. I. Sec. 8 Cl. 8 of the Constitution, by ensuring inventors are granted appropriate protection for their works.[16] In February 2024, the USPTO responded to the October 2023 Executive Order by setting out a guideline for examiners and applicants to use when there is integration of generative AI involved.[17] The guideline calls for a consider of the factors found in the ‘significant contribution test’ set out in Pannu v. Iolab Corp.[18] The test intends to verify if the inventor, a natural person, made significant contributions in the process.[19]

Inventor

The phrasing and use of the word inventor in the Constitution does not explicitly define who or what an inventor is.[20] However, under 35 U.S. Code §100 the law defines inventor as the individual, or individuals who invented or discovered the subject matter of the invention.[21] The USPTO has followed the interpretation that inventors relate to human beings, by denying patent applications that were derived by AI.[22] In the application for patent protection, some individuals listed AI as the inventor and named a natural person as the owner of the patent (assignee).[23]

However, in 2022, the Federal Circuit determined that only a natural person can be an inventor because 35 U.S.C. 100(f) defined the word ‘inventor’ as individual or individuals. [24]

Generative AI

Generative AI (Gen AI) has the ability to produce human-like work.[25] It can also analyze, describe, and use algorithms, prompts, or images to produce novel work.[26] It has presented various benefits such as improved efficiency and clarity in claim writing as well as lower costs of research for companies and the USPTO.[27]

Particularly within the pharmaceutical industry, companies have implemented Gen AI to spark innovation in the research and development field.[28] In this field, researchers try to decode protein structure folding and develop new molecules and compounds that enable discoveries of new pharma products.[29] Gen AI tries to find predictive patterns for these structures and presents a visual representation of how the protein’s pattern works, which allows scientists to narrow their research to specific areas.[30] This technology implicates generating new drug designs with reduced time and costs of development.[31]

Significant Contributions

Significant contributions have been an issue for joint inventors.[32] Mostly because there was dispute as to who qualified as an inventor when participating in the creation of a product.[33] In 1998, under Pannu, the Court held that a joint inventor is someone who: (1) has contributed in some significant manner to the conception or reduction to practice of the invention, (2) contributes to the claimed invention that is not insignificantly in quality, when measured against the dimension of the full invention, and (3) does no more than merely explain to the real inventors’ well-known concepts and/or the current state of the art.[34]

The factors above are now being utilized to determine the true inventor in patent applications using Gen AI.[35] Here, for a person to be granted the status of inventor, they must satisfy all three prongs.[36] Under the first prong, the guideline suggested that mere input into a Gen AI system is not a significant contribution.[37] The proposed inventor must show how they utilized the information produced by the Gen AI to develop the invention.[38] Next, an inventor must leverage the full scope of the invention with their personal contribution and assert these contributions are of quality.[39] Lastly, the inventor must disclose specific knowledge that influenced the invention and is not general to others.[40]

III. Discussion

The implementation of these guidelines, in response to a Presidential Administration Executive Order, provides a clearer view of what lies ahead.[41] While machines are not inventors, they certainly aid inventors in their research and analysis.[42] Using these Gen AI systems provides companies with the most ideal and scarce benefit: time.[43] Having more time, companies can enjoy exclusive market rights and help reduce patent over-fillings. Thus, spurring innovation and competition among inventors while maintaining the proper implementation of these systems.

The guideline sets out that the use of these Gen AI systems in the future will be key to how courts will navigate lawsuits and how examiners will review patent applications.[44] The USPTO has made it clear that the inventor must be a human and not a machine or any technology.[45] However, it did not rule out using AI in making a novel subject matter discovery.[46] The guideline explains that when using Gen AI, the most important factor to look at is whether the inventor has made any significant contribution.[47] This standard stems from the significant contribution test found in Pannu.[48] The USPTO emphasizes that simply entering information into the Gen AI system is not a significant contribution. In addition, the Pannu test would be applied to each claim to verify that each has a significant contributor that is human.[49]

On the other hand, although these factors provide helpful guidance on how to approach patent applications that used AI in the process of invention, there are still other factors lingering. For example, upon discovery one party may request that the private chat communications between the inventor and AI be submitted to trial. This can cause data privacy issues on what information can be presented in court and could affect the business operations of the company or any of its projects it plans to patent.

In addition, this evidence could be crucial in determining how significantly a person contributed to the invention. Therefore, the balancing of private chat communications with inventorship guidelines can be both critical to determining the level of the inventor’s contribution and harmful to the inventor’s future projects.

An additional question the legal field may need to navigate is whether the agency and the courts should consider what rights the inventor of the Gen AI system has. The court should consider whether the creator of these systems has relieved any authorship of what is generated by AI systems or if the terms and conditions of AI implementation provides any insight.

IV. Conclusion

It is clear that the legal field has been slow in adopting new sets of rules that govern the complex issues of today. Particularly, because the topic is so novel, there is a need for understanding how broad or narrow these proposed rules and laws should be. The legal field wants to cover as much as it can but not be restrictive as to hinder the use of technology.

Agency guidelines are not set in stone and are not the law. While the guideline provided by the agency is helpful, it still leaves many other questions unanswered. Such as: would the creator of the Gen AI be entitled to any inventorship or royalty? How much can be disclosed in trial when using Gen AI? Are litigation tactics allowed to be more aggressive in discovery?

Thus, there needs to be more prompt guidance and statutes governing these issues. However, at the same time, it needs to be considered that the guideline came as a response to an Executive Order. Which begs the question, would there have been any guidance if not for the Order? In my opinion, the answer is no. The agency might have waited to see how the court system would address this issue and go from there. However, based on the guideline the courts had addressed identifying who qualifies as an inventor in patent applications back in 1998 when the federal court decided Pannu. Thus, there needs to be more involvement from the USPTO in establishing future guidelines and identify potential hurdles. As soon as these agencies and courts tackle these underlying issues it can be clearer as what is to come.

[1] Andrei Iancu and Rama Elluru, When AI Helps Generate Inventions, Who Is the Inventor?, Center for Strategic and International Studies (Feb. 22, 2014), https://www.csis.org/analysis/when-ai-helps-generate-inventions-who-inventor [https://perma.cc/7VRX-2DXJ].

[2] Id.

[3] U.S. Const. art. I, § 8, cl. 8.

[4] Iancu and Elluru, supra note 1.

[5] Id.

[6] U.S. Const. art. I, § 8, cl. 8.

[7] Kathy Vidal, AI and inventorship guidance: Incentivizing human ingenuity and investment in AI-assisted inventions, United States Patent & Trademark Office (Feb. 12, 2024), https://www.uspto.gov/blog/ai-and-inventorship-guidance-incentivizing [https://perma.cc/9XD3-WK59].

[8] Id.

[9] What is generative AI?, McKinsey & Company, (Apr. 2, 2024), https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-generative-ai [https://perma.cc/C5N3-K2HV].

[10] Vidal, supra note 7.

[11] Exec. Order No. 14,110, 88 Fed. Reg. 75191 (Nov. 1, 2023).

[12] Id.

[13] USPTO issues inventorship guidance and examples for AI-assisted inventors, United States Patent and Trademark (Feb. 12, 2024), https://www.uspto.gov/subscription-center/2024/uspto-issues-inventorship-guidance-and-examples-ai-assisted-inventions [https://perma.cc/DD8P-4NVN].

[14] Id.

[15] About us, United States Patent & Trademark Office, https://www.uspto.gov/about-us [https://perma.cc/L9VB-YFQX].

[16] Id.

[17] Vidal, supra note 7.

[18] Id.

[19] Id.

[20] U.S. Const. art. I, § 8, cl. 8.

[21] 35 U.S.C. § 100.

[22] Annie Bolton et al., Can Machines Invent Drugs: Patent Law seeks to Prioritize Human Innovation as Artificial Intelligence Inventions Rise, JD Supra (Apr. 30, 2024), https://www.jdsupra.com/legalnews/can-machines-invent-drugs-patent-law-9225455 [https://perma.cc/9RKR-TNKU].

[23] Id.

[24] Id.

[25] What is generative AI?, supra note 9.

[26] Bolton, supra note 18.

[27] Id.

[28] Carrie Arnold. Inside the nascent industry of AI-designed drugs, Nature (June 1, 2023), https://www.nature.com/articles/s41591-023-02361-0 [https://perma.cc/4FRA-LTQC].

[29] Id.

[30] Alexandre Blanco-Gonzales, et al. The Role of AI in Drug Discovery: Challenges, Opportunities, and Strategies, National Center for Biotechnology Information (June 1, 2023), https://pmc.ncbi.nlm.nih.gov/articles/PMC10302890/ [https://perma.cc/2A35-9SKH].

[31] Id.

[32] Pannu v. Iolab Corp., 155 F.3d 1344 (Fed. Cir. 1998).

[33] Id.

[34] Id at 1351.

[35] Vidal, supra note 7.

[36] Id.

[37] Pannu, 155 F.3d at 1344.

[38] Id.

[39] Id.

[40] Id.

[41] Exec. Order No. 14,110.

[42] Blanco-Gonzales, supra note 30.