Fifty-One Laboratories for AV Law: A Comparative Study of State AV Legislation, 2017–2026

• In Event: AI Policy, Law, and Regulatory Diffusion

Abstract

From driverless taxis and commercial freight trucks to sidewalk delivery robots, autonomous vehicles are entering shared public space in multiple forms and at an accelerating pace. State legislatures have become the primary arena in which the terms of that entry are being worked out. Drawing on a systematic comparative analysis of state-level AV legislation introduced between 2017 and 2026, the paper examines what governance approaches states are taking, how those approaches travel from one jurisdiction to another, and what governance gaps these early governance efforts have left open. The study examines pending, passed, and rejected legislation across the fifty states and the District of Columbia along eight governance dimensions covering authorization scope, preemption architecture, liability assignment, and classificatory reach, in order to show how legislative templates are diffusing across states and how states are positioning themselves along a spectrum that runs from promotional to precautionary. From this coding the paper develops a governance-posture typology that sorts jurisdictions into seven categories, including enabling, multi-layered, precautionary, institutional, stalemate, and legislative silence, and tracks each jurisdiction along one of four posture trajectories (stable, contested, evolving, or stalled) that capture how governance approaches change over multiple legislative sessions. The analysis gives particular attention to the durability of early enabling frameworks and the conditions under which subsequent sessions attempt to revisit or restrict particularly early permissive frameworks, a backlash cycle that appears in nine states and typically unfolds within one to seven years of the original enactment. Additionally, because AV governance has been built almost entirely on top of existing vehicle-code categories, it leaves significant classificatory blind spots around autonomous devices that operate outside the traditional road network, with implications for how pedestrian spaces, sidewalks, and other shared environments are governed as automation spreads. The paper contributes to literatures on risk governance, policy diffusion, and institutional change, and it offers state and local policymakers a practical framework for anticipating where current regulatory approaches are likely to fall short as autonomous vehicle form factors continue to diverge from the car-on-road paradigm that shaped the first generation of AV law.

Author

• Lina Moe, Rutgers University
  Presenting Author