Applying the panel's User Decision Framework, the case studies offer a step-by-step look at how agencies and practitioners can navigate concrete AI decisions
FOR IMMEDIATE RELEASE
September 16, 2026
Contact: Brian Edsall
bedsall@counciloncj.org
(202) 793-3373 ext 204
WASHINGTON – The Council on Criminal Justice (CCJ) Task Force on Artificial Intelligence today released three case studies illustrating how criminal justice agencies and communities can apply its User Decision Framework to evaluate, implement, and oversee different AI tools in policing, public defense, and corrections.
The Task Force released the framework in March 2026, providing specific, detailed steps agencies should take as they decide whether and how to implement AI tools in justice operations.
The case studies apply that framework to three hypothetical scenarios grounded in real-world AI tool categories already being used or considered across the field. Each case study follows the framework’s five decision phases, from assessing organizational readiness and weighing risks and opportunities through procurement, implementation, and ongoing management and reassessment.
“The same AI tool can look very different depending on who’s using it and what’s at stake. As the pace of AI development accelerates, criminal justice agencies can’t afford to implement first and ask questions later,” said Task Force Chair Nathan Hecht, a former chief justice of the Texas Supreme Court. “These case studies show how agencies can use our framework to weigh the tradeoffs, decide whether to move forward, and establish the safeguards and protocols needed to use AI effectively and responsibly.”
The three case studies explore:
- Police report writing: Examines use of AI tools that draft incident reports from body-worn camera audio and other inputs, focusing on accuracy, risks of overreliance on AI-generated drafts, and preservation of AI prompts and drafts for legal scrutiny.
- Public defense: Examines use of general-purpose AI tools for research, drafting, document review, and case preparation, including the confidentiality protections, attorney verification, and office policies needed to reduce workloads without compromising the quality of representation.
- State corrections: Examines use of AI-powered tools that support classification decisions, including how to assess accuracy and disparities, preserve meaningful human review, and allow incarcerated people to challenge their classification levels.
Together, the case studies demonstrate how the framework translates into practical decisions about vendor contracts, staff training, pilot testing, ongoing oversight, and more. Each scenario considers how AI use compares to existing practices and realistic alternatives, including non-AI approaches, and identifies circumstances in which agencies should reconsider or avoid use.
“Artificial intelligence is one of the most polarizing topics in criminal justice today, but it doesn’t have to be,” said Adam Gelb, CCJ President and CEO. “It’s not the technology alone that determines whether AI helps or harms. It’s also how and why it’s used, who is accountable when it doesn’t work as intended, and how performance is measured and improved over time. These case studies give criminal justice leaders concrete examples to work from so they don’t have to reinvent the wheel with every new tool that inevitably comes along.”
The case studies draw on information from practitioners and technology vendors, publicly available research and product documentation, and the insights and expertise of Task Force members. They are not intended to endorse, reject, or rate any specific vendor, product, or technology category in the abstract. Instead, they show how agencies and practitioners can use the framework’s step-by-step structure to reason through concrete AI questions, taking into account the proposed use, affected population, legal obligations, and available oversight capacity.
About the CCJ Task Force on Artificial Intelligence
Launched in June 2025, the CCJ Task Force on Artificial Intelligence is a national, nonpartisan initiative to develop standards and evidence-based recommendations to guide the effective and ethical use of AI in the criminal justice system. The Task Force released a set of guiding principles in October 2025.
The Task Force’s 15 members represent AI technology developers and researchers, police executives and other criminal justice practitioners, civil rights advocates, community leaders, and formerly incarcerated people. Its work is supported by CCJ staff and researchers at RAND, a leading research organization with extensive expertise in criminal justice and emerging technologies.
Support for the Task Force comes from the Georgia Power Foundation, Heising-Simons Foundation, The Just Trust, Microsoft, and The Tow Foundation, as well as the John D. and Catherine T. MacArthur Foundation and other CCJ general operating contributors.
About the Council on Criminal Justice
The Council on Criminal Justice is a nonpartisan think tank and invitational membership organization that advances understanding of the criminal justice policy choices facing the nation and builds consensus for solutions that enhance safety and justice for all.
The Council does not take policy positions. As part of its array of activities, the Council conducts research and convenes independent commissions composed of Council members who produce reports with findings and policy recommendations on matters of concern. Commission findings and conclusions are not subject to approval of the Council’s Board of Directors, Board of Trustees, or funders. See more information on the Council’s policy projects here.
