American judges, particularly Federal District Court Judges, are possibly the most professionally well-documented people in the world. Every opinion, every motion, every three-minute hearing is meticulously recorded and provided for public dissemination via PACER. It is this massive dataset that has allowed for the development of modern analytics services like Lex Machina, Ravel Law, Westlaw and Bloomberg Analytics, and so forth. However, the functionality of these services has always been constrained by the logistical issues with assigning new metadata as search/sorting criteria: there’s simply too much data to classify, so Analytics services have always relied on the metadata assigned by PACER or whatever state e-filing system they draw from.

AI eliminates these logistical constraints. It can classify the data by any number of new, non-obvious criteria outlined by its trainer, allowing analytics services to wield the exhaustive record of a judge’s career to reveal preferences and proclivities in a way that may shake our profession to its core.