Lawyers Beware: LawGeex is Here

March 26, 2018 – Artificial intelligence (AI) is being applied to the review of business contracts by an Israeli company, LawGeex, and the software in competition with lawyers is showing them up. The latest test involves the analysis of 5 non-disclosure agreements (NDAs). NDAs are among the most common company contracts created by in-house lawyers to ensure confidentiality when intellectual property and other corporate secrets get shared. So choosing NDAs to show how AI could outperform lawyers makes sense.

So what was the test?

The 5 NDAs contained 153 paragraphs each and featured complex technical and legal jargon.

Twenty U.S. trained lawyers with decades of contract review experience went up against the software.

LawGeex proved far more accurate than the average level of accuracy among the 20 lawyers. The AI achieved 94%. The lawyers 85%. The most accurate human matched LawGeex at 94%. The least accurate scored 67%.

Of more significance is the amount of time LawGeex needed to its analysis: 26 seconds.

The average for the 20 lawyers was 92 minutes with the shortest time 51 and the longest 156.

Adjudicating the test were Stanford, Duke, and University of Southern California law school professors and the Computer Science department of Bar-Ilan University in Tel Aviv.

States Gillian Hadfield, Professor of Law and Economics at Southern California, “The lawyers who reviewed these documents were fully focused on the task: it didn’t sink to the bottom of a to-do list, it didn’t get rushed through while waiting for a plane or with one eye on the clock to get out the door to pick up the kids. The margin of efficiency is likely to be even greater than the results shown here.”

So how did LawGeex do it?

LawGeex was founded in 2014 by Noory Bechor, an international lawyer, and Ilan Admon, an AI expert.

The algorithms these two created were tested on tens of thousands of NDAs.

The AI tool combines machine learning, text analysis, and natural language processing all processed within the cloud.

It incorporates best practices from top legal minds in the United States.

When presented with a contract, it reads and analyzes it in microseconds, then suggests changes based on the known pre-defined legal policies of a company.

It doesn’t require contracts to be fed to it in a specific format because it can scan and read a wide range of electronic document types from .doc to .docx to PDF to JPEGs.

When it encounters discrepancies between a submitted contract and company policy it redlines the problems and alerts the lawyers involved to fix them.

It provides fast turnaround. The time savings amount to 80% when compared to standard contract review.

LawGeex captures and updates institutional legal knowledge in the process of reviewing contracts. That knowledge is retained in a commonly shared knowledge base.

Should lawyers be worried about their jobs?

According to Hadfield, LawGeex, and AI tools like it, solve two problems. They make contract turnaround much faster and more reliable, and they free up human legal resources to focus on other company issues.

A full report of the test results can be accessed online that describes the methodology, analysis, and results, and includes insights from leading legal scholars on the study’s long-term impact and the role of AI in the future of law.