Addressing Capacity Constraints in Corrections Through Agentic Artificial Intelligence

• In Event: Bridging the Gap: Artificial intelligence for Smarter Decision-Making in Criminal Justice

Abstract

Corrections agencies face persistent challenges in adopting data-driven decision-making, including outdated information systems, limited technical capacity, and chronic workforce shortages. Many continue to rely on professional judgment rather than data-informed insights—not from a lack of interest, but due to insufficient tools, training, and infrastructure to effectively leverage data. Without strong analytic capacity, agencies struggle to allocate resources efficiently and implement measurable improvements. This paper presents the development of an agentic artificial intelligence (AI) system designed to support decision-making in corrections. Unlike traditional approaches, agentic AI operates with a degree of autonomy, enabling it to initiate actions, interpret complex environments, and adapt to human input. Especially in an environment where skilled analysts are difficult to recruit and retain, agentic AI offers a promising solution to bridge critical gaps in knowledge and capacity. Using prison population forecasting as a use case, this paper contributes to the growing field of intelligent system design for justice applications, providing a pathway toward more effective, transparent, and cost-efficient decision-making through AI technologies.

Author

• Kideuk Kim, Urban Institute