Indicium ex Machina: Unstructured Sentencing and Disparate Outcomes in Canada

Abstract

Recent advancements in artificial intelligence have caused a wave of technological normality. As expected, the criminal justice system, and more increasingly, criminal sentencing is seeing a trend of “technosolutionism” due to real concerns about unjustified disparity. Truly, artificial intelligence has the prospect of making the sentencing process more effective, value-driven, consistent, and predictable. However, relying on the assumption that using such a system may require being confined to the normative sentencing traditions of each country, this thesis argues that there are crucial questions to be addressed about how this technological normality fits within the normative pillars of extant legal principles, especially in an anomalous sentencing jurisdiction like Canada. Despite sufficient incentives to integrate AI, the lack of a meaningful sentencing structure significantly undermines the prospect of AI mitigating disparity. To effectively harness the potential of an automated system, the current sentencing approach must substantially shift direction towards a well-structured sentencing practice.