Scaling Success: Takeaways from the ACA Winter Conference on AI, Education, and Reentry
Earlier this month, I had the privilege of speaking at the American Correctional Association’s Winter Conference in Long Beach. Joining a panel of speakers from across the country, we dove into a conversation that is top of mind for many: how can technology become a true bridge for better reentry outcomes?
In the collaborative learning session that followed the panel, we heard from the practitioners in the room–all tasked with the challenge of finding new approaches to rehabilitation. Four specific insights seemed to resonate most deeply during that conversation.
1. AI as the "Overeager New Recruit"
One theme that immediately struck a chord was the role of AI in reducing the administrative load on staff. I shared a metaphor of AI as an "overeager new recruit": incredibly hardworking and capable of tackling the mountain of tedious tasks that currently bury case managers—like synthesizing years of case notes or verifying housing availability—yet requiring a seasoned human "supervisor" to ensure quality and equity.
For leaders dealing with staffing shortages and burnout, the prospect of automating paperwork isn't just about saving minutes; it’s about giving case managers the headspace to build the human relationships that are the backbone of successful reentry. We’re seeing this in the correctional centers using our Case Planning Assistant across Utah. Instead of running through a rote list of questions, staff are going into conversations with a suggested list of priorities for each person and finalizing those plans together in real time. The AI assists with understanding and matching needs, but it is the human beings in the room who decide the final path forward.
2. Moving the Goalposts: Measuring What Matters
For too long, corrections has been measured by a single, negative metric: the recidivism rate. While it’s an important data point, it’s a lagging indicator. It also doesn't tell the story of a person’s growth or how the people who don’t return are contributing to their communities.
Mark Spahr from the Maine Department of Corrections pointed out that while systems know that things like housing and salary are the real indicators of a person's stability, they have historically lacked the data infrastructure to track them.
At Recidiviz, we are helping partners such as the Nebraska Department of Correctional Services build these quantitative feedback loops, measuring program participation and positive outcomes through AI-powered case note analyses, and evaluating the relationship between them. We’re then going further and encouraging incarcerated people to engage in proven programming through new features like Colorado’s program catalog, available to all residents through the Opportunities app on their tablets.
3. Personalization in Case Planning and Education
The challenge of providing individualized attention within a massive system is a perennial one for corrections leaders. The shift toward personalization is now coming to both case planning and education simultaneously.
In the room, we explored how AI helps match individuals with the specific resources they need for a successful transition while also transforming education by tailoring learning pathways to individual learners. Whether it’s through Nebraska’s reentry checklist or personalized learning modules, the leaders recognized that putting this information directly into the hands of residents increases their agency and motivation in a way that traditional, static plans simply cannot.
4. Bridging the Continuity Gap
Finally, Dr. Brooke Wheeler from North Carolina highlighted a critical barrier to rehabilitation: the "continuity gap" between different agencies and facilities. For leaders like Brooke, the goal is ensuring that a person's educational and vocational progress isn't lost when they move through the system.
Technology is becoming the connective tissue that allows a learner's pathway to follow them. By creating a unified, data-driven record of progress, we can ensure that when someone transitions from one facility to another—or eventually back to the community—their roadmap remains intact and actionable.
Looking Ahead
As panelist Allen Wachendorfer noted, the best approach to technology is to start with the problem, not the solution. By focusing on the specific barriers in reentry and education, we can ensure that AI is a force for good.
The work being done by our state partners is proof that when we combine human expertise and local resource networks with smart technology, we can build a system that is more transparent, more effective, and ultimately, more human.
