Automating Discretionary Decision-Making in the Public Sector: Legal Dimensions

Abstract

Complex legal issues accompany proposals to automate the exercise of public sector functions. Deterministic and predictive (data-drive) algorithmic decision-making systems challenge the function of legal principles which are deeply-entrenched in Anglo-phone legal systems. Public law protects human dignity by enforcing standards of rationality, transparency and responsiveness on public officials. This paper discusses the legal risks which arise where traditional public law doctrines confront algorithmic government.

Keywords: algorithmic decision-making, public law, artificial intelligence, machine learning, human dignity, administrative law, Australia

JEL Classification: K00, K42