Thaler and Contextually Sound AI Regulation, in Recreating Creativity, Reinventing Inventiveness: AI and Intellectual Property Law (Nikos Koutras & Niloufer Selvadurai eds., 2026)
Files
Description
Artificial intelligence decision-making poses a thorny regulatory problem, largely because theorists and regulators assume that AI plays the same role in human legal systems that humans do. The general assumption is that the legal context for human regulation is the same as the legal context for AI regulation. For example, asking the question: “Is an AI an inventor?" only makes sense if one assumes that an AI “inventor”, whatever that means, can be fitted into the legal framework intended for human inventors, should be rewarded with a state-granted monopoly, can meaningfully be sanctioned for bad behaviour, and so on. We argue they cannot. This chapter theorises, describes, and advocates for a general system of artificial intelligence regulation based on the idea that human incentive-generating institutions like profit, sanction, liability, ex ante explainability and the like fail to influence AI’s decision-making towards human welfare. Rather, we propose, the straightforward solution is to reward and sanction humans who develop and deploy AI for the damages or benefits those deployed products bring to other humans. This chapter will use Thaler v. Commissioner of Patents [2021] FCA 879 and related cases as grounding examples. Our focus is not on the specific outcomes of the cases but rather on the problems for regulation of AI that the cases highlight. We propose that existing legislation may not be able to address the challenges posed by AI technology and thus advance a concrete set of policy moves to create a general system of artificial intelligence regulation. Fortunately, our policy intervention requires no great changes: humans are responsible for the AI they deploy, and those systems of reward and sanction are presently in place. Our approach contains a methodological critique as well. The Thaler cases are emblematic of a series of attempts to backdoor AI into legal personhood through an impoverished view of the meaning of words. If legal words derive meaning solely through their definitions, then statements like “an AI can be an ‘inventor’" have a certain surface plausibility. We note that words draw their meaning from an entire community and context of meaning, and that more attention to how legal terms gain meaning will prevent courts from stumbling into profoundly inapt legal regulatory regimes for AI.
ISBN
9781032196282
Publication Date
2026
Publisher
Routledge
Disciplines
Computer Law | Intellectual Property Law | Internet Law | Law | Science and Technology Law
Repository Citation
Nikos Koutras & Joshua Fairfield, Thaler and Contextually Sound AI Regulation, in Recreating Creativity, Reinventing Inventiveness: AI and Intellectual Property Law (Nikos Koutras & Niloufer Selvadurai eds., 2026),
https://scholarlycommons.law.wlu.edu/fac_books/222