Predictive Justice: Examining the Role of Risk Assessments in Pretrial Decisions

• In Event: Sentencing and AI

Abstract

The use of risk assessments in pretrial custody determinations has expanded significantly, often eclipsing other legal considerations. Originally intended as just one factor in bail decisions, risk has become a central focus, shaping judicial outcomes in ways that remain highly contested. Methods for assessing risk vary widely, from judicial intuition to actuarial tools and more modern algorithmic models. Supporters claim that these tools enhance accuracy in predicting court appearances and public safety risks while reducing unnecessary pretrial detention. However, critics argue that there is little evidence to support these claims, pointing to systemic biases in algorithmic decision-making and the potential erosion of statutory safeguards. The increasing reliance on predictive risk assessments raises critical questions about their role in reinforcing punitive logics, their impact on pretrial justice, and whether they truly serve the goals of fairness and efficiency.

Authors

• Kelly Hannah-Moffat, University of Toronto (Canada)

• Fernando Avila, Brock University