AI inventors: deference for legal personality without respect for innovation?

Abstract

Thirty years ago, Lawrence B. Solum wrote a visionary article concerning the debate on the legal personhood of artificial intelligence (AI) and posed the following question: Could an artificial intelligence become a legal person? As of today, this question is only theoretical.1 Thirty years later, in the area of patent law, recent cases from the UK, the USA and Australia have resoundingly answered: ‘no’ to intellectual property rights (IPRs) for AI systems. While academic commentators have perhaps helped sensationalize the concept of an ‘AI inventor’,2 national patent offices and court judgments have taken a sober approach in opining that AI systems cannot possess or be subject to IPRs in patent law. This article argues that recent decisions, which have declined to grant IPRs to AI systems, are correct and pay deference to the requirement of legal personality, which is the foundation of law. Granting legal personhood to AI systems or machines, at least within the area of intellectual property, seems implausible. However, the article further argues that the deference given to legal personality in AI inventor cases raises the question of what impact this has on innovation. An approach that outcasts AI systems from patents may well respect the current order of legal personhood within the boundaries of most legal systems, but it may negatively impact the existing incentives for innovation provided by the IPR system. The article advocates for a reform of intellectual property laws to take into account the role of human creativity in enabling the creativity of AI systems.