Artificial intelligence and the "natural person" criterion in patent inventorship

As generative artificial intelligence (AI) systems become integral to scientific and technological innovation, courts and patent offices are grappling with whether, and how, AI assistance affects inventorship. Current doctrine in leading jurisdictions limits inventorship to a "natural person", yet the accelerating convergence of robotics, synthetic biology, and neurotechnology exposes gaps in statutory definitions and policies. This article challenges the legal requirement of "natural person" for patent inventorship. It also discusses litigation involving AI‑named inventors. This article also covers thought experiments involving legally recognized non‑human or non‑biological persons ( e.g ., rivers granted legal personality, robots with formal citizenship) and biologically modified entities, to test boundary conditions and highlight policy risks. The discussion in this article revolves around the lack of clarity with cases of hybrids that could potentially fulfill the criteria of "natural person" and the operational mechanics of patent law including conception, reduction to practice, and chain of title and assignment. The article ends with suggestions related to clarifying the statutory language related to "natural person" inventorship. Without timely reform, inconsistent interpretations across jurisdictions could jeopardize patent validity, distort incentives, and invite strategic abuse. To address this, the article proposes clarifying the definition of "natural person" in patent law to eliminate interpretive ambiguity and expressly exclude artificial and hybrid entities from inventorship. The "natural person" requirement for patent inventorship warrants clarification in the age of artificial intelligence.