AI Predicts Judge Rulings With 86% Accuracy

AI company’s software predicts judge rulings by taking into account their history & biographical data.

A new tool with a database of the history of judges’ rulings and biographical information is predicting potential rulings and helping lawyers better decide how to invest their time and resources. Plaintiffs are also seeking out favorable courts and judges based on the tool’s data, which makes the whole science more precise.

The database is developed by a startup called Pre/Dicta and includes all state and federal civil litigation cases as well as the biographical information of the judges, like the city they were born in. It’s not currently intended to predict the outcomes of criminal cases and jury trials. Looking at all the data, the tool can predict patterns and identify potential areas of bias.

There are a total of 120 data points per judge including where the judge went to law school, what their net worth is, how they rule when the lawyers are from big law firms versus boutique practices, and the judges’ history in public law, private practice and state judgeships, reports Axios.

The CEO told Axios that the tool can predict how a judge will rule with 86% accuracy without even considering the facts of the case.

It’s now possible to get a good understanding of how a judge is likely to rule in a case without even feeding the tool any data on the specifics of what the case is about. As a growing industry and science, predicting case outcomes can significantly change the way plaintiffs and lawyers use their money and time.

It’s notable that court observers can already predict, to some extent, how likely the judge is to rule in favor of one side. This is closer to 50%, however, and Pre/Dicta’s tool has been operating at 86% accuracy for a year.

The CEO says that the product may be close to hitting an accuracy ceiling.

By Abhimanyu

Unwrapping the fast-evolving AI popular culture.