Predictive Coding: Revolutionary or a Fading Trend?

As members of the legal community continue to strive to provide the best possible value and service to their clients, solo practitioners and international law firms alike have recently begun turning to predictive coding in an attempt to increase productivity while decreasing the costs associated with document review and production. Although the science behind predictive coding—also commonly known as computer-assisted review, computer-based advanced analytics, technology-assisted review, or predictive tagging—has been around since the late 2000s, it has only been within the past two years that predictive coding has begun to gain nationwide acceptance from practicing attorneys and the court.

Predictive coding refers to software technology that is trained to identify and determine the relevancy of a large number of documents. In short, predictive coding allows computers to search voluminous quantities of documents while utilizing complex algorithms to identify relevant and privileged documents in place of a human document reviewer. When used properly, predictive coding can significantly decrease the amount of time and costs associated with human review. However, predictive coding software has not cut out the human reviewer entirely. Rather, attorneys continue to play a significant role in this phase of the litigation as they are required to collect documents, de-duplicate documents, determine what constitutes a relevant and/or privileged document, and most importantly, train the software to accurately determine whether a document is relevant and/or privileged.