AI, Legal Pluralism, and Property Governance: Comparative Insights on Rulemaking and Enforcement from the U.S., U.K., Ukraine, China, and India

Abstract

Artificial intelligence (AI) systems are set to assume a growing number of tasks that have been the traditional domain of human rulemaking and rule-enforcing agencies. However, the world into which AI will be deployed is one of legal pluralism and hybrid property governance. Building on legal pluralist scholarship and on parallel developments in the US, UK, Ukraine, China, and India, this article provides an exploratory analysis of how AI systems may fit with plural property rulemaking and enforcement regimes that encompass both formal law and informal social norms. We also use Elinor Ostrom’s Institutional Analysis and Development (IAD) framework to analyze how AI systems may operate in a plural legal environment where local, community-specific, informal “rules-in-use” may depart from stated, formal “rules in books” to produce a system of hybrid property governance. AI has the potential to bring higher levels of efficiency and consistency to such administrative tasks. At the same time, we find that if AI systems are designed to ignore the social and legal pluralism in which they are embedded, they may well erode public trust, legitimacy, and justice in highly socially complex contexts that are too variable or local to be treated as standardized or to have rule-of-law principles uniformly imposed on them. We therefore argue that the operational design details of AI systems and their use in hybrid governance arrangements matter, that rule enforcement algorithms that are context-blind or context-oblivious are likely to have distributive impacts that increase conflict and injustice, and that the context matters because local governance arrangements do. Cases of socially contextualized AI property governance systems, from automated traffic cameras in India to predictive policing in the UK to mortgage fraud detection in the US, illustrate a tension between a desire to automate the standardized enforcement of rules using AI and people’s desire for relational social norms. The article presents a framework and some concrete design considerations to help guide the participatory design of AI in plural property governance contexts that surfaces, engages, and accounts for stakeholders, local norms, and legitimacy criteria. In so doing, we aim to contribute to and expand the normative and institutional AI governance literature as well as the literatures on legal pluralism and institutional design.