The burgeoning field of artificial intelligence is revolutionizing drug discovery, with companies like Insilico Medicine leveraging sophisticated AI platforms to generate novel drug candidates at an unprecedented pace. Insilico, for instance, recently announced a promising drug for pulmonary fibrosis, enthusiastically proclaiming in a press release that the molecule was "discovered by" its generative AI platform. This marks a significant advancement, as AI models are now capable of designing complex molecular structures with a speed and ingenuity that can surpass human capabilities, potentially accelerating the development of life-saving therapies. However, this technological leap has illuminated a critical and complex intersection with intellectual property law, particularly concerning patent rights and inventorship. When Insilico sought to protect its groundbreaking chemical structure through a patent, the company conspicuously omitted any mention of its AI. Instead, the patent application listed five human individuals, including CEO Alex Zhavoronkov, as the drug’s "inventors." This deliberate omission highlights a fundamental tenet of current intellectual property law: credit for invention, and thus the right to legal protection, is exclusively reserved for humans.

The legal landscape surrounding AI inventorship has been shaped by a series of landmark cases and judicial interpretations. A pivotal moment occurred when Ryan Abbott, a partner at the law firm Brown, Neri, Smith & Khan, initiated a pro bono test case. Abbott sought to name an AI system, dubbed DABUS, as the inventor of a novel food container. The container’s design, characterized by an intricate geometric surface that enhances heat transfer and stackability, was entirely conceived by the AI without direct human input. Abbott argued that if an AI generated an invention, it should be recognized as the inventor, thereby raising profound philosophical questions about AI rights and the very definition of innovation. However, in 2022, an appeals court in Washington, D.C., dismissed these "metaphysical matters." The court’s reasoning was grounded in the plain language of U.S. statutes, which define an inventor as an "individual," a term understood to mean a human being. Consequently, machines, lacking personhood, cannot legally be recognized as inventors. This ruling, in essence, closed the door on AI as an inventor under current U.S. law.

The implications of this legal precedent are far-reaching for the rapidly evolving biotech sector. Sarah Korman, patent attorney and Chief Business Officer and Legal Officer at Isomorphic Labs, an Alphabet spinout focused on AI-driven drug discovery, articulated this challenge. She stated at MIT Technology Review’s EmTech event that "there needs to be a human inventor or there’s no invention and no patent." Korman also acknowledged that existing laws are not adequately equipped to handle the pace of AI development and will require significant evolution. The U.S. Patent and Trademark Office itself has recognized that AI systems, much like other tools, can perform actions that, if performed by a human, would constitute inventorship. This suggests that while current law may preclude AI from being named an inventor, the AI’s contribution to the inventive process is undeniable.

The central challenge moving forward, therefore, is not whether AI can invent, but rather whether any human has contributed sufficiently to an AI-generated discovery to be legitimately named as an inventor. Abbott warns of potential legal challenges to AI-developed drugs, as misidentification of inventors can serve as grounds for invalidating a patent. This concern is amplified by the U.S. Copyright Office’s stance on AI-generated works. The office has refused to grant copyrights to images and text created by AI, a decision that has raised alarms among organizations like the Motion Picture Association of America, whose members are increasingly utilizing these generative tools.

The fundamental purpose of intellectual property law, as enshrined in Article 1 of the U.S. Constitution, is to foster innovation by granting inventors and authors exclusive rights to their creations for a limited time, thereby "to promote the Progress of Science and the Useful Arts." The current approach by the U.S. patent office appears to be a "don’t ask, don’t tell" policy regarding the use of AI in invention. Under the Biden administration, guidance was issued to help applicants navigate the complexities of human inventorship in AI-assisted discoveries. However, this approach shifted under the Trump administration, with the patent office now characterizing AI as merely a tool, akin to a calculator, whose use need not be disclosed.

In light of these legal ambiguities, pioneering AI drug companies are diligently ensuring human involvement in their processes, at least for the time being, and meticulously documenting every step. Alex Zhavoronkov of Insilico Medicine explained that human chemists are still essential for synthesizing the drug candidates, developing variations, and conducting animal testing. "That’s the person who is going to be named on the patent," Zhavoronkov stated. He further elaborated that even if the entire process, including experiments, were to be fully automated, a human would still be responsible for initiating the process and allocating resources. This raises a provocative question: does the act of pushing a button or approving a budget constitute inventorship? Abbott suggests this is a matter for future legal consideration. He posed a hypothetical scenario: "What if I asked Claude to cure cancer, and it did? I think it would be inappropriate to claim that I invented that." This highlights the evolving ethical and legal debates surrounding AI’s role in creation and the future of intellectual property in an era of increasingly sophisticated artificial intelligence. The journey from AI-generated molecular design to a patented, life-saving drug is now intricately intertwined with the human element, a balance that legal frameworks are still struggling to define.