Book Title: Computational Criminology: AI Applications in Forensic Science and Criminal Justice
Editors: Dr. Xavier Louis, Dr. Surbhi Girdhar, Ms. Aswathi Chandran Nair, Mr. Ravi Kumar, and Ms. Nandini Katare
Chapter: 22
DOI: https://doi.org/10.59646/704/22
Author: Akash Chauhan
Abstract
The pretrial detention decision to detain an accused person prior to trial is one of the most impactful exercises of state power in the criminal justice process, and it is made in highly pressurized, information-poor and structurally unequal environments. In several jurisdictions, algorithmic risk assessment tools have been adopted as aids to decision-making to minimize arbitrary and disproportionate detention decisions. This chapter discusses the organisation and design of pre-trial risk assessment tools, the impact of pre-trial risk assessment tools on detention rates and racial disparities, the constitutional and due process considerations of the use of pre-trial risk assessment tools, and transparency and accountability demands of responsible deployment. It considers risk scores as tools of advice, not classification, and compares the experiences of New Jersey and Kentucky in pretrial reform as large-scale natural experiments to assess the potential impact of algorithmic pretrial assessment on system injustice; and charts the ongoing debate about the potential for algorithmic pretrial assessment to increase or decrease systemic injustice.