close
close

topicnews · October 24, 2024

Research Report: AI to Scale Legal Reform: Mapping and Redacting Racial Agreements in Santa Clara County

Research Report: AI to Scale Legal Reform: Mapping and Redacting Racial Agreements in Santa Clara County

U.S. property deeds often contain discriminatory language that excludes people of certain races from buying a home or specifies that only whites are allowed to do so. Even though the Supreme Court has ruled that such racially restrictive covenants are unenforceable, they still litter deeds across the country and continue to be signed by homeowners when a property changes hands.

In 2021, California passed a law requiring the state’s 58 counties to create programs to identify and redact deed records containing racial covenants.

Peter Henderson, an assistant professor in the School of Public and International Affairs and its computer science department at Princeton, worked with researchers at Stanford University to curate a collection of racial alliances from various jurisdictions around the country and train a state-of-the-art open language model to create racial alliances with high accuracy recognize.

As a pilot application of the technology, the researchers worked with Santa Clara County, California, which needed to review more than 84 million pages of deed documents, providing a good test case for how to reduce the manual effort of this task. The system identified more than 7,500 racial agreements, which are confirmed and processed by district attorneys.

“It’s an example of how open source models can be used for the public good,” Henderson said. “Homeowners do not have to sign such offensive terms much sooner than a purely manual review would have allowed.”