A Métis prisoner challenged risk assessments influencing correctional decisions. The majority found that CSC knew of longstanding concerns about the instruments' cultural validity but had not taken reasonable steps to investigate their accuracy when used for Indigenous offenders.
The Court declared a breach of section 24(1) of the Corrections and Conditional Release Act. It did not find a Charter section 7 or 15 infringement on the record. The holding concerns a statutory duty to validate information before using it in decisions that shape a person's custody and rehabilitation.
The contested instruments were used to classify the risk posed by a prisoner and to inform correctional programming and release decisions. Ewert argued that the tools were developed and tested principally on populations unlike Indigenous offenders and that the Correctional Service of Canada knew of significant concerns about their cross cultural validity. The majority did not hold that every actuarial tool was necessarily inaccurate for Indigenous offenders. It held that the prison authority had failed to take reasonable steps to investigate whether those tools produced reliable information for this group.
Section 24(1) of the Corrections and Conditional Release Act required reasonable steps to ensure information about offenders was accurate, current and complete. The Court granted a declaration of statutory breach but did not find that the evidence established Charter violations under sections 7 or 15. The case has wider significance wherever institutions use statistical predictions to justify restrictions on individuals. It illustrates why a risk model cannot be assumed reliable for a population merely because it is used routinely, particularly where the state knows of serious validation questions.