ADAPPT Lab

Our work

We help policymakers make informed decisions in jurisdictions across the United States.

Below, we highlight several projects where our innovations in computational policy reshaped policymaking and resulted in lasting change, along with a slate of new efforts we are advancing now.

Lady justice with a black redaction bar covering her eyes serving as a blindfold.

Race-blind charging

We developed a generative artificial intelligence algorithm that automatically redacts race-related information from police reports, allowing prosecutors reviewing these redacted documents to make race-blind charging decisions.

In collaboration with the Computational Policy Lab.

A hand holding a smartphone displaying a text message court date reminder.

Reducing jail incarceration with behavioral nudges

In a partnership with a major public defender’s office, we found that automated court date reminders reduce jail involvement for missed court dates by 20%—the first time this benefit has been scientifically demonstrated.

In collaboration with the Computational Policy Lab.

Illustration of a police traffic stop at night.

Analyzing police stops

We developed new approaches to understand police stop practices, leading to novel insights, new datasets, and policy impact.

In collaboration with Big Local News, the Computational Policy Lab, and the Policing Project.

Emerging projects

We are looking for public sector partners and philanthropic support for these new projects. If you're interested, reach out!

Reducing SNAP payment error rates

Helping states avoid cost-sharing penalties with AI that helps applicants submit materials, automatically checks documents, and flags cases for human review.

In collaboration with Karen Chatfield at the Bank Street College of Education.

Tackling the digital evidence explosion

Building generative AI tools that help public defenders and prosecutors handle an avalanche of digital evidence, reducing incarceration and improving client representation.

A national database of police misconduct

Using generative AI to extract and organize information from federal civil rights cases against law enforcement agencies, resulting in a structured database for any researcher to use.

In collaboration with Greg Stoddard and the Policing Project.

Computational tools for child services

Helping child welfare agencies manage complex caseloads and pull valuable information from case notes to improve outcomes for children.

Quantifying perceptions of public safety

Using computer vision to test longstanding theories on the connection between street conditions and perceptions of public safety.

Catalyzing diversion of 911 calls

Creating a shared evaluation platform for 911 call technology, with benchmarks and datasets that will help governments and vendors assess promising improvements.

In collaboration with the Government Performance Lab.

Algorithms, Data, and Public Policy Technology (ADAPPT) Lab

Based at NYU PRIISM.
Supported by:
All content © 2026 ADAPPT Lab. All rights reserved.