Negligent AIs and the pursuit of a new model of personhood

Artificial intelligence (or ‘AI’) personhoods have generated a host of legal and philosophical views, and there exists a wealth of opinions, both positive and negative, on granting that status. Many disagreements arising between scholars seem to be rooted in their initial starting points, wherein scholars adhere to different positions on the bases of personhood bestowals, which are themselves controversial and do not align easily where conflicts arise. As a result of this conundrum, the work regarding AI personhoods tends to be fractured into ideological camps that are splintered in accordance with the biases that each scholar holds, which makes consensus a difficult prospect. However, this is a challenge not just for AI personhoods, but for personhood overall, and the arrival of AIs as the newest contenders for personhoods offers a chance to reconsider how personhoods are granted. The prospect of an AI’s personhood highlights many of the issues in the assumptions that are made when granting personhoods to humans and nonhuman ‘artificial’ persons, like the appeals to physical humanity, a need for key facets, and personhood’s largely unquestioned connection with rights. Using case law and legislation, along with legal and philosophical literature on AI and other personhoods, this thesis will explore the hesitancy in the law to grant AI personhoods and argues in favour of such a status while using AIs as a case study to explore the debates in personhood discourse at present. This AI personhood status will then be applied in a practical manner, through the lens of one specific legal field in which that status could be applied. The selected field for this thesis is the tort of negligence, rooted within the broader field of private and tort law, which will be justified throughout this thesis on the basis of a range of factors that set negligence apart as an appropriate field in its theoretical and practical ambits for personhood. The outcome of that review will be a new model for the bestowal of personhoods, one that will be deployed to consider not only the facets that a person may possess but also the reasons why an entity is being personified to determine what capacities and vulnerabilities that status should entail. This approach is inspired by Max Weber, who analysed the reasons why humans act (or legislate) in social situations, and his works serve as the model’s starting point. The model will be applied to negligence, a subject within Common private law, to assess its effectiveness. Under this model, the instrumental and normative reasons for an entity’s personification will be used to make decisions, along with claims about the entity’s physical, mental, and metaphysical aspects. This variety of factors will be used to arrive at a more nuanced, synthesised position than the current regime offers, which will impact not just new candidates like AIs, but also on other nonpersons that have long been contenders for personhoods (like animals), as well as humans who are currently considered persons by default.