Home BlogDataset Building A Public Database to Train Legal AI

Building A Public Database to Train Legal AI

by Michael McLaughlin
Harvard Law School.

The Harvard Law School Library has published a database of 6.4 million court cases to spur the development of AI applications dealing with the law. The library scanned over 40 million pages of court decisions, and the data includes every official report for cases that occured in the United States from 1658 to June 2018. While there have been some prominent examples of AI used for legal applications, such as analyzing data in record-intensive investigations, the limited availability of training data has been an obstacle to its widespread adoption.

Get the data.

Image: Chensiyuan.

You may also like

Show Buttons
Hide Buttons