What happened
RAND, in collaboration with the Council on Criminal Justice Task Force on Artificial Intelligence, published a comprehensive taxonomy of AI applications across policing, courts, corrections, and community supervision. The report addresses a critical gap: AI tools in criminal justice—from facial recognition and automated police report writing to case scheduling, classification, and violence prediction—are often discussed and governed as if they were a single category of technology, when in fact they differ widely in technical sophistication, intended use, and influence on human judgment. The taxonomy groups applications by the functions and decision points they support, illuminating how similar technologies may serve different purposes across settings and providing a structured framework for stakeholders responsible for evaluating, procuring, and implementing AI in justice systems.
Why it matters
Criminal justice AI decisions affect public safety and individual rights at scale. This taxonomy provides the first structured framework for distinguishing high-risk from lower-risk applications, enabling policymakers and system leaders to evaluate whether specific AI tools are appropriate, effective, transparent, and equitable—rather than applying blanket governance approaches that either over-restrict low-risk tools or under-govern high-stakes deployments.
Action needed
State justice system CIOs and policy officials should map their current and planned AI deployments against the taxonomy by Q3 to identify governance gaps and prioritize oversight resources.