Automating government decision-making: implications for the rule of law

This chapter assesses the benefits and challenges to the rule of law posed by automation of government decision-making. In this regard, reference should be made to a few short commentaries which call for more attention to be paid to the governmental context: see, for example, Mikhaylov, Esteve, and Campion 2018; Kennedy 2017; and Perry 2017. It adopts the rule of law as a standard because it is accepted worldwide as providing normative guidance on the appropriate conduct and operation of governments. The rule of law is a ubiquitous, yet elusive concept, at the heart of which lies a widely held conviction that society should be governed by law. However, in this chapter, our goal is not to provide yet another account of the rule of law (modern accounts include Lord Bingham 2007: 69; Tamanaha 2004: 2; and Gowder 2016). Instead, we critically investigate how principles of the rule of law are affected by the increasing use of two kinds of automation: human-authored pre-programmed rules (such as expert systems) and tools that derive rules from historic data to make inferences or predictions (often using machine learning).