Peter Bo Zhang (University of Toronto, Faculty of Law) has posted Constructive Scienter: An Animal-Law Answer to the AI Responsibility Gap, Law, Innovation and Technology (forthcoming 2027), on SSRN. Here is the abstract:
This article recovers an overlooked common law doctrine, the scienter action’s heightened liability for the keeper of a dangerous animal, and brings it to bear on the opaque algorithmic systems through which the state now makes consequential decisions about individuals. It argues that the less a public body can explain a coercive decision, the more it must justify it. Where public power, algorithmic opacity, and an affected person’s right to reasons converge, the system’s opacity can neither discharge that duty nor excuse its evasion. Under the scienter action, knowledge of a creature’s propensity strengthens rather than severs the keeper’s responsibility. Carried into public law, that logic yields a new attributional concept, ‘constructive scienter’, the deployer’s knowledge of what it cannot know. So framed, the much-discussed ‘responsibility gap’ is, in public law, no gap at all. The dominant response of AI legal personhood misreads the problem, conflating the exclusion that personhood may remedy in animal law with the evasion it would enable in AI governance. The argument runs through State v Loomis, with implications for administrative decision-making and judicial review.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum