Predictive Coding Cleared By SDNY; May Signal a Win For More Effective and Efficient Document Review

In a recent order, U.S. District Judge Andrew Carter, Jr. upheld Magistrate Judge Andrew Peck’s ruling that predictive coding be employed in a case involving, among other things, review of some three million e-mails. Predictive coding is a software-based review methodology that allows lawyers to “train” a computer program to recognize relevant keywords and other textual features in documents. In short, the lawyer reviews a sample set of documents, marking certain features (keywords or others) as relevant, and feeds this information to the software, which then learns to use the lawyer’s criteria to filter out irrelevant materials.

New Scientist has a great summary of the case. As far as the access to justice community is concerned, Judge Carter’s order may represent a milestone in the interaction between technology and the judiciary. Manual document review by humans is a time-honored fixture of the legal profession, but it isn’t particularly efficient, may create prohibitive costs for litigants, and can, in certain cases, be less effective than technological review.

Thomas Gricks, defense counsel in this case, claims that predictive coding would allow him to conclude his document review in two weeks, and at only one percent of the cost of a thorough manual review. Technology that reduces costs and increases effectiveness and efficiency when properly applied has tremendous applications for enhancing access to justice. Technology-enhanced, efficient discovery processes allow lawyers to take more clients, each at reduced costs. The rise of this sort of streamlining should be of great interest to all access-to-justice advocates.