chevron-down Created with Sketch Beta.
April 29, 2024

Why AI Developed Drugs May Not Develop Into Patents

By Leslie R. Snider

The intersection between artificial intelligence (AI) and life sciences promises to be a thriving path producing many benefits for the pharmaceutical industry, especially in drug discovery and development. Per the U.S. Food and Drug Administration (FDA), AI offers numerous possibilities for the pharmaceutical industry to bolster continuous improvement though optimizing process and control, smart monitoring and maintenance, trend monitoring, and more. With these advancements, in recent years, the FDA has seen a rapid increase in the number of drug applications utilizing AI and machine learning (ML), most pertinently in the stages of clinical research and development (R&D). It is also typically during this R&D phase when most companies apply for a patent with the U.S. Patent and Trademark Office (USPTO) seeking patent protection of their new drug. However, recent gray areas in the law have called the patentability of these AI-developed drugs and biologicals into question, leaving the door open for resolution between tradition versus innovation. 

Current Market for AI and Drug Development

Recent business forecasts have reported that “[t]he global AI market was valued at nearly $59.7 billion in 2021, and is estimated to reach $422.4 billion by 2028.” Specific to the pharmaceutical industry, recent data shows that AI startup companies in the business of drug development generated approximately $2.1 billion in the first half of 2021 alone. Additionally, recent data shows that the U.S. is the global leader in participating with and hosting such AI startup drug development companies with about 55.10%, followed by Europe with about 19.90% of such companies.

FDA Embraces Novel AI Tools for Drug Development

Even outside of the business market, the FDA has made recent investments and advancements in the use of AI in drug development. In 2020, the FDA launched its Innovative Science and Technology Approaches for New Drugs (ISTAND) pilot program aiming to support the development of novel drug development tools (DDTs) that fall outside of the traditional existing parameters for DDT qualifications to encourage novel approaches to drug development for regulatory approval. According to the FDA, examples of these DDTs can include tools that may help enable remote or decentralized trials, ​tools that may advance the understanding of drugs, or tools that leverage digital health technologies (DHTs). In January 2024, the FDA accepted to the ISTAND program its first AI-based and DHT-based submission, which generates an automated depression and anxiety severity measurement tool. It accomplishes this by utilizing AI/ML and DHTs such as computing platforms, connectivity, software, and/or sensors to support drug development and review. Per the FDA, this acceptance is a “pioneering step” that “…aligns with FDA’s vision of optimizing drug development and evaluation, potentially expediting the availability of safe and effective treatments.”

AI Patents Filings and Grants

According to the USPTO, patent applications containing AI consisted of about 9% of the applications in 2002, compared to about 16% of applications by 2018. Recent data also shows that the global pharmaceutical industry saw a 52% increase in AI-related patent applications during the second quarter of 2023 as compared to quarter one. However, the number of actual grants of patents only rose by 15% between the two quarters. Despite this disparity, given the booming business market for AI in drug development, the recent acceptance by the FDA of an AI-based submission for DDT qualification, and the ever-increasing AI-related patent filings in the pharmaceutical industry, it is clear that AI’s involvement in drug development is an intertwined partnership that is here to stay. However, as the USPTO highlighted in 2021:

“A final area of increasing policy interest is the impact of recent Supreme Court jurisprudence on the patent eligibility of AI technologies. A series of Supreme Court decisions, starting in 2010 with Bilski v. Kappos, have drastically altered the types of inventions that are eligible for patent protection. These decisions generally reduced the patent eligibility of inventions that contain abstract ideas, laws of natures [sic] and products of nature….” (emphasis added).

Although recent trends are showing the multitude of benefits of utilizing AI/ML in the drug discovery and development process, the path to obtaining a patent for such drug may not be as clear.

Necessity for Human Inventorship

In 2022, in Thaler v. Vidal, the U.S. Court of Appeals for the Federal Circuit was faced with the question of whether an AI program may be listed as the sole inventor on a patent application. Relying solely on the plain meaning of the term “inventor” in the Patent Act, the Federal Circuit affirmed that “an inventor must be a natural person” and “must be human beings.” The Supreme Court subsequently denied certiorari in 2023. In light of this, it is clear that an AI program cannot be listed as the sole inventor on a patent application. However, the question still lingers as to whether inventions made by human beings with the assistance of AI are eligible for patent protection. Can an AI program be listed as a joint inventor on a patent application?

The Patent Act defines the term “inventor” as the “individual or, if a joint invention, the individuals collectively who invented or discovered the subject matter of the invention.” The Patent Act proceeds to define the terms “joint inventor” and “coinventor” as “any [one] of the individuals who invented or discovered the subject matter of a joint invention.” As explained, Thaler holds that the inventor must be an individual, and an “individual” must be a human being. Since joint inventors must also be individuals, it is also likely that joint inventors must also be human beings. This is consistent with other Federal Circuit decisions that hold only a natural person is capable of “perform[ing] [the] mental act” that is the “touchstone of inventorship.” Therefore, if an AI program cannot be named as a sole or joint inventor on a patent application, how much AI is too much so that human inventorship, a clear necessity for patent eligibility, is not impeded?

A Pressing Need for Clarity

Post-Thaler, the intersection between AI and patent eligibility raises many questions, especially as the FDA sees an unprecedented increase in the number of submissions of drugs and biologicals utilizing AI/ML during development. According to recent data, in 2016 and 2017, the FDA only received one submission of drugs and biologicals utilizing AI/ML each year. Now, those numbers have skyrocketed exponentially, with the FDA receiving approximately 132 drug and biological submissions utilizing AI/ML in 2021, and approximately 170 submissions in 2022. In February 2023, the USPTO issued a notice to the Federal Register highlighting the Thaler decision when soliciting comments from stakeholders regarding the “uncertainty” around AI inventorship as AI/ML makes “greater contributions to innovation.” However, in October 2023, the White House got involved.

The Executive Order, titled Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, sets definitive deadlines for federal agencies, including the USPTO, to impose new actions and issue new guidance on their incorporation of AI into their respective fields. Per Executive Order 14110, the USPTO had until February 27, 2024, to publish guidance addressing inventorship and the use of AI in the inventive process, which the USPTO published on February 13, 2024, and until July 26, 2024, to issue additional guidance on other AI considerations, including updated guidance on patent eligibility.

In its February 2024 guidance, the USPTO directly addresses Thaler and instructs that AI-assisted inventions are not unpatentable due to improper inventorship; however, both a natural person and the AI system must make a “significant contribution” to the invention to be considered inventors. The USPTO concedes that determining the significance of a natural person’s contribution to an AI-assisted invention may be difficult to ascertain, but provides a list of principles for consideration influenced by the factors in Pannu v. Iolab Corp. as there is no bright-line test.

Finding Neutral Ground

As implied by the USPTO, the neutral ground between AI/ML and human inventorship for the patent eligibility may lie in finding balance between how significantly each contributes to the invention. Pannu provides several determinative factors to establish significant contribution, such as the inventor (1) contributing in a significant manner to the conception or reduction to practice of the invention, (2) contributing to the claimed invention that is not insignificant in quality when measured against the dimension of the full invention, and (3) doing more than merely explaining to the real inventors well-known concepts and/or the current state of the art. Per the USPTO, failure to meet any one of these factors precludes the ability to be named as an inventor. As seen in Thaler, the AI program was named as the sole inventor of the invention without a human also being named as an inventor, and the Federal Circuit prohibited the issuance of a patent.

Although there is remaining uncertainty surrounding patent eligibility of AI-developed products, even in light of the February 2024 USPTO guidance, there is hope on the horizon for those wishing to obtain such a patent. In October 2022, Lantern Pharma, an AI-based company, announced its issuance of a new patent for their drug candidate that utilized AI/ML “to transform the cost, pace, and timeline of oncology drug discovery and development” that is targeted for “never-smokers with relapsed non-small cell lung cancer.” Again in August 2023, Lantern Pharma announced its notice of allowance from the USPTO of another drug candidate, used for treatment of atypical teratoid rhabdoid tumors, that utilized AI/ML “to advance to a first-in-human Phase 1 basket trial.”

Either way, the drug development community needs answers from the USPTO to achieve equilibrium between AI/ML and human inventorship for patent eligibility. Regardless of how these inventorship issues are resolved, the benefits produced from the intersection with AI/ML are likely to help many, especially with advancements in pharmaceuticals and life sciences. It helps that the USPTO has now issued its long-awaited guidance on AI-assisted inventions, with its focus on human contributions and ingenuity. The sooner there is clarity in this area, the clearer the incentives will be for inventors—and their employers—working in this cutting-edge arena.

Leslie R. Snider

U.S. Department of Health and Human Services, Washington, DC

Leslie R. Snider is an attorney with the U.S. Department of Health and Human Services located in Washington D.C. She is an inaugural fellow of the ABA Health Law Section Diversity & Inclusion Fellowship Program. She can be reached at [email protected]

    The material in all ABA publications is copyrighted and may be reprinted by permission only. Request reprint permission here.