Machine learning predictive algorithms and the policing of future crimes: governance and oversight

Oswald, Marion and Babuta, Alexander (2020) Machine learning predictive algorithms and the policing of future crimes: governance and oversight. In: Policing and Artificial Intelligence. Routledge, Abingdon, Oxon ; New York, NY. (In Press)

[img] Text
SSRN-id3479081.pdf - Accepted Version
Restricted to Repository staff only

Download (320kB) | Request a copy

Abstract

This chapter focuses upon machine learning algorithms within police decision-making in England and Wales, specifically in relation to predictive analytics. It first reviews the state of the art regarding the implementation of algorithmic tools underpinned by machine learning to aid police decision-making, and notes the impact of austerity as a driver for the development of such tools. We discuss how what could be called ‘Austerity AI’ is often linked to the prevention and public protection common law duties and functions of the police, a broad and imprecise legal base that the ECtHR in Catt found less than satisfactory. The potential implications of these tools for appropriate application of discretion within policing, as well as their potential impact on individual rights are then considered. Finally, existing and recommended governance and oversight processes, including those designed to facilitate trials of emerging technologies, are reviewed, and proposals made for statutory clarification of policing functions and duties, thus providing a clearer framework against which proposals for new AI development can be assessed.

Item Type: Book Section
Uncontrolled Keywords: Machine learning, algorithms, policing, discretion, Austerity AI
Subjects: G700 Artificial Intelligence
M200 Law by Topic
Department: Faculties > Business and Law > Northumbria Law School
Depositing User: Elena Carlaw
Date Deposited: 05 Nov 2019 09:37
Last Modified: 05 Nov 2019 09:45
URI: http://nrl.northumbria.ac.uk/id/eprint/41361

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics