Background
In State v Loomis, the defendant challenged his sentence on due-process grounds after the court relied on a COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) risk-assessment algorithm. Loomis argued the proprietary system’s secret methodology prevented him from contesting its accuracy and violated his constitutional right to an individualized sentence. The Wisconsin Supreme Court upheld the sentence, finding no due-process violation, but imposed cautionary limits on the future use of such tools.
AI interaction
‘A COMPAS risk assessment should not be used to determine whether an offender is incarcerated or to determine the severity of the sentence.’ The Court accepted algorithmic assessments as permissible supplementary information but warned that opaque, proprietary scoring could not replace judicial discretion. The decision is significant as one of the first high-level rulings to address AI transparency, fairness, and accountability in criminal justice, shaping later debates on algorithmic bias and explainability in sentencing.
Note: State v Loomis remains a cornerstone precedent on the constitutional limits of algorithmic decision-making in sentencing. It highlights how courts may allow AI-based tools while warning against their determinative or unexamined use in liberty-impacting decisions.