Title: Bureaucratic bastardry: robodebt/debt recovery, AI and the stigmatisation of citizens by machines and systems

Authors: Adam Graycar; Adam B. Masters

Addresses: The Stretton Institute, The University of Adelaide, G29 Napier Building, North Terrace Campus, 5005, Australia ' Transnational Research Institute on Corruption, The Australian National University, RSSS Building, 146 Ellery Crescent, Acton, 2601, Australia

Abstract: Automation in public administration is inevitable and can bring great benefits. Seventy years ago, Isaac Asimov foresaw the need to protect humans from their creations in his novel I, Robot, drafting three laws to ensure robots: 1) did not injure a person, or allow them to be harmed; 2) obeyed orders, that did not conflict with the first law; 3) protected their own existence without compromising the first or second laws. The first two laws can apply to the automated systems created by government to administer public service. This paper examines the failed policy to automate welfare debt collection in Australia. Despite repeated warnings the malign policy caused considerable harm to clients and resulted in a billion-dollar settlement against the government. Using a framework centred on the concepts of organisational evil and bureaucratic animosity, we suggest that such complex undertakings could benefit from reference to Asimov's laws.

Keywords: malign policy; robodebt; public values; administrative evil; bureaucratic bastardry.

DOI: 10.1504/IJPP.2022.127432

International Journal of Public Policy, 2022 Vol.16 No.5/6, pp.333 - 344

Received: 14 Sep 2021
Accepted: 04 Mar 2022

Published online: 05 Dec 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article