Administrative Law, Artificial Intelligence, and Procedural Rights

Abstract

As artificial intelligence (AI) advances its capacity to replicate human reasoning, governments increasingly rely on AI to support and supplant human decision-makers. This evolution is already discernible in many Canadian administrative agencies, where AI is being implemented to support legal determinations with automated prediction and pattern recognition. Algorithmic tools have the potential to optimize our legal systems with increased efficiency, consistency, and precision. However, as automated systems begin to shape legal decisions, we can also expect a shift in how the law is interpreted and applied. This paper examines how such a shift will impact procedural fairness. In particular, this paper analyzes how AI-driven decisions threaten the right to an unbiased decision-maker and the right to receive reasons for a decision. After outlining the unique threats posed by different types of AI, I propose three recommendations for policy reform. Ultimately, I argue that new regulatory safeguards are required to better protect procedural fairness and government accountability.

Keywords: Administrative law; artificial intelligence; algorithmic bias; procedural fairness