US patent law requires human inventors for AI-generated drug designs despite company claims

The event highlights a contradiction between how companies market AI-driven discoveries and how they secure legal protections. Insilico Medicine’s press release framed its AI platform as the primary discoverer of a pulmonary fibrosis drug, yet its patent application adhered to US legal requirements by naming only human inventors. This dual messaging reflects the tension between technological capability and regulatory frameworks that have not evolved to accommodate AI’s role in innovation.

US patent law currently restricts inventorship to humans, a stance reinforced by court rulings like the DABUS case. The appeals court dismissed arguments about AI’s creative contributions, emphasizing that statutes define inventors as “individuals,” a term interpreted as human. This legal precedent forces companies to attribute AI-generated inventions to human contributors, even if their role is minimal, such as approving budgets or initiating computational processes.

The discrepancy raises practical risks for AI-driven biotech. Patents can be invalidated if the listed inventors are incorrect, and AI-generated drugs could face legal challenges if courts determine no human contributed sufficiently to qualify as an inventor. This uncertainty may deter investment in AI drug discovery, as companies weigh the benefits of speed and novelty against the potential for weakened intellectual property protections.

The US Patent and Trademark Office’s current guidance treats AI as a tool, requiring no disclosure of its use in patent applications. This “don’t-ask-don’t-tell” approach allows companies to navigate the system but does not address the underlying legal ambiguity. Future litigation or legislative changes may be needed to clarify inventorship standards, particularly as AI systems become more autonomous in generating patentable discoveries.

For engineers and biotech firms, the event underscores the importance of documenting human involvement in AI-driven processes. While AI can propose novel drug designs, legal protections still depend on demonstrating human contributions, such as synthesis, testing, or decision-making. Until laws evolve, companies will need to balance AI’s capabilities with the need to satisfy patent office requirements.