Predictive coding, also known as automated review, represents an evolving technology that provides litigants with a computer-assisted alternative to manual review of large document sets. Although numerous implementations of the technology exist, nearly all involve a team of attorneys manually reviewing a small sampling of documents to “train” the computer how to code the larger pool of remaining documents.

Recently, a number of opinions have been issued permitting, or even encouraging, the use of predictive coding. As a result, some pundits and jurists have gone so far as to suggest that search terms are dead, or

at best, an antiquated method for culling documents. Yet, while the use of predictive coding can reduce the cost of review, it is not a magic bullet or a good fit for every case. Indeed, in many cases, search terms paired with a well-managed human review will still provide the most cost-effective option without sacrificing accuracy.

This alert highlights developing trends in predictive coding, explores the range of associated consequences and suggests appropriate scenarios for using predictive coding in lieu of other forms of computer-assisted review.

Please see full alert below for more information.