Authors: Dominik Stammbach, Peter Henderson
In a speech at the NJ AI Summit on April 11, 2024, Phil Murphy, then-Governor of New Jersey, announced his support for a new partnership between Princeton University and New Jersey’s Office of the Public Defender (NJOPD). “Together, they are working to build an AI-powered resource that will help our state’s public defenders provide the best possible representation to our most vulnerable neighbors.“
Together with the NJOPD, we developed the OPD Resource Library, an AI-powered search tool to surface relevant information within the office. Given a legal question, the tool searches through appellate briefs, internal documents, and public directives, and returns the most relevant ones. The library makes the collective knowledge of the office accessible to all employees within seconds.
The NJ AI Task Force Report summarizes the intended use case as follows: “[The Resource Library] is an AI-powered tool designed to help public defenders efficiently draft legal briefs. The tool is a repository of legal briefs, motions, and other legal documents that public defenders can use as templates or references to streamline the drafting process.” The library is inspired by current office practices, where defenders are encouraged to consult past briefs, “facilitating more efficient case preparation and promoting consistency in legal arguments” (New Jersey Lawyer).
Speaking at the OECD Roundtable on Equal Access to Justice (2025), Jennifer Sellitti, the NJ public defender, described the tool as a significant time saver: “Public defenders can now enter a legal question and receive content vetted by NJOPD experts. Our lawyers still exercise full professional judgment and verify all results, but the tool eliminates the most tedious part of brief writing. What once took hours – gathering the core legal arguments – now takes seconds.”
Our collaboration provides a blueprint for how AI can improve public services and empower public servants—if developed responsibly in close collaboration with public partners. As part of our effort, we worked closely with NJOPD to identify best practices for AI-based deployments for public defense. To make the resource as accurate as possible, we trained customized AI models and stress tested these in multiple evaluation rounds. With help from the NJ Innovation Authority, New Jersey Office of Information Technology, AWS engineers and the NJOPD IT team, we deployed and launched the tool as a closed universe application, accessible to all NJOPD attorneys and staff.
We also conducted structured qualitative interviews to help inform best practices and potential blue sky ideas for how to leverage AI for public defense. Defenders across the country emphasized that the most helpful AI applications for legal research would be to provide high-level summaries, information overviews, and starting points for finding the best facts and ideas for their case. These are the same reasons why consulting past briefs has long been common practice within the office.
Our work also furthered artificial intelligence research. We showed that off-the-shelf AI systems don’t always work well for underexplored use cases like public defense retrieval, and release an associated benchmark. By leveraging the combined on-the-ground expertise of public defenders and AI expertise at Princeton, we can build and deploy better, responsible AI systems for public good, and advance corresponding research.
At the OECD Roundtable, Sellitti further argued: “Technology will not define the future of public defense. Our lawyers will.” In our interviews, one participant said that “clients have a right to counsel, not to machines.” We believe that AI for public defense should follow such humanistic principles, and we design our AI tools accordingly.
For further reading, we point to:
- The NJOPD Resource Library Launch
- A more extensive blog post on our broader research agenda around AI and public defense
- A preprint Legal Retrieval for Public Defenders, outlining our approach to training AI models for responsible deployment in public defense.
- A manuscript about Public Defenders’ Perspectives on AI Adoption, FAccT 2026
We would like to thank all our partners at the New Jersey Office of the Public Defender, most importantly Jennifer Perez, Alison Perrone, Brandon Rios and Ronald Wildmann, without whom this would not have been possible. We also thank Walker Gosrich, who has supported the project from the start, anonymous interview participants who took part in our qualitative research, and everybody from the NJ Innovation Authority, New Jersey Office of Information Technology and AWS.
Leave a Reply