Algorithmic Decision-making and Legality: Public Law Dimensions

Abstract

Automating the exercise of statutory powers through algorithmic decision-making carries high levels of legal risk. Fundamental public law doctrines assume that legal powers will be exercised by a particular kind of decision-making agent: one with sufficient cognitive capacities to understand the interpretative complexity of legal instruments and respond to highly dynamic environments. Public law doctrines also assume that clear reasons can be given for the exercise of public power and, by default, attribute legal responsibility for the exercise of statutory powers to a human being bearing political and social responsibility. Those doctrines provide the standards against which the legality of algorithmic decision-making in the public sector must be tested and until they are met lawyers should be very sceptical of suggestions that statutory powers can be automated.