Year: 2026
Type: conference-paper
Abstract: Scholarship on artificial personhood has largely asked whether AI systems should hold rights, conducted in the shadow of unresolved metaphysical debates about machine consciousness. This paper proposes a different entry point. I draw on tools from the study of institutional path dependence and evolutionary population dynamics (Arthur, 1994; Pierson, 2000; Price, 1970), on recent work by Chilson and Schwitzgebel (2026) treating AI as strange intelligence, and on the legal tradition governing vulnerable persons in England and Wales: the Mental Capacity Act 2005, the Court of Protection, and the human rights architecture of the UN Convention on the Rights of Persons with Disabilities. Following Bartlett (2023), I argue that the Mental Capacity Act’s trajectory illustrates the risk of value lock-in at a formative stage of governance development — the tendency of early conceptual choices to become entrenched and resistant to revision even after the conditions that produced them have changed. I contrast this with Convention-aligned reforms in Ireland and Peru, whose implementation records show how difficult escaping lock-in actually is. I then apply path dependence and evolutionary population dynamics to present-day governance decisions about AI systems and digital minds. The paper makes an architectural claim about governance under uncertainty. It does not claim moral equivalence between humans and AI systems.
DOI: 10.1162/isal.a.995
Author: Angie Normandale
Institution: University of York
Language: English
Subfield: Law
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7213329777