Case Study: How Measures for Justice Is Building the Foundation for Responsible AI in Criminal Justice
Situation
Artificial intelligence is transforming how organizations access and use information. In criminal justice, it is a must that technology be paired with transparency, accuracy, and trust.
For more than a decade, Measures for Justice has worked with prosecutors, courts, law enforcement agencies, and communities to improve the quality, accessibility, and transparency of criminal justice data. Through tools like Commons, MFJ helps jurisdictions make complex justice data easier to understand and use.
As AI capabilities continue to evolve, MFJ saw an opportunity to address a critical challenge: making trusted criminal justice data easier for people to access while ensuring that humans remain in control of decision-making.
The future of AI in criminal justice is not about replacing expertise, it is about helping people find, understand, and use reliable information more effectively.
Problem
Criminal justice agencies and communities increasingly need access to timely, accurate data to answer important questions about system performance, equity, and outcomes.
While platforms like Commons make high-quality justice data publicly available, finding the right information can still be challenging. Users often need to navigate multiple filters, measures, and visualizations to locate the insights they need.
This creates several challenges:
- Practitioners may struggle to quickly find relevant data to inform decisions.
- Community members may face barriers when trying to understand how their justice system operates.
- Users need confidence that AI tools will provide accurate, verifiable information.
- Traditional AI approaches can introduce risks, including inaccurate summaries, unclear sourcing, or conclusions that users cannot independently verify.
In criminal justice, where data can influence policies and decisions that affect people’s lives, accessibility cannot come at the expense of accuracy or trust.
Solution
Measures for Justice has been exploring how responsible AI can make trusted criminal justice data easier to access without compromising transparency or human judgment.
Rather than asking AI to interpret criminal justice data or make decisions, MFJ is using AI as a navigation tool for helping users find the information they need while allowing them to draw their own conclusions.
Through a partnership supported by Microsoft, Measures for Justice developed the Commons AIÂ Navigator, a conversational tool designed to help users locate relevant visualizations within the Commons platform using natural language.
Instead of requiring users to know which filters, measures, or dashboards to select, they can simply ask a question, such as:
“How do prosecutor decisions vary by race?”
The AI Navigator then identifies the most relevant Commons dashboards and visualizations, guiding users directly to the underlying data.
The tool does not summarize findings, interpret results, or generate recommendations. Instead, it serves as an intelligent guide that improves access to trusted information while preserving transparency and user control.
The Commons AI Navigator is currently being piloted with the Yolo County District Attorney’s Office and the Multi-Cultural Community Council through funding and Azure AI resources provided by Microsoft.