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: 21
DOI: https://doi.org/10.59646/704/21
Author: Mr. Naresh A
Abstract
Actuarial risk assessment instruments (RAIs) statistical tools that assign individuals to risk categories predicting recidivism likelihood have been deployed throughout the United States and internationally as decision-support aids in sentencing, parole, and probation. Proponents argue that structured actuarial assessment reduces judicial arbitrariness and anchors decisions in empirically validated risk factors; critics contend that such instruments encode structural inequalities, violate due process by penalising factors beyond an individual’s control, and generate racially disparate outcomes. This chapter traces the development of RAIs from early clinical prediction instruments through validated actuarial tools such as COMPAS, LSI-R, PSA, and Static-99, examines the technical methodologies of their construction, evaluates the competing statistical fairness criteria invoked in their assessment, and analyses the legal and constitutional challenges that their use has generated. The chapter argues that the deployment of RAIs in high-stakes criminal justice decisions demands a higher standard of transparency, validation, and procedural safeguard than current practice achieves.