In recent months, three different courts have considered the use of computer assisted review of documents to select relevant materials for production in connection with litigation. The decisions by two of these courts approving the use of computer-assisted review technology are indicative of an overall trend in both state and federal courts towards greater automation of the discovery process. Given the vast amounts of electronic data currently generated by businesses of all sizes, we expect this trend to continue and the implications of these decisions to be significant. Therefore, knowledge of available computer-assisted review technology and the changing scope and magnitude of discovery of electronically stored information is vital for any organization that may find itself faced with a demand for the production of information from an adversary in a litigation context.

What Is Computer-Assisted Document Review?

Computer-assisted review of documents entails the use of computer software that utilizes an algorithm which enables it to learn which documents in a subset are relevant to the issues being litigated and which are not. The software acquires its learning through interaction with a human reviewer. Unlike manual review—which is typically conducted by more junior lawyers —computer-assisted review requires a senior associate or partner, i.e., someone with extensive familiarity with the case and the legal issues involved, to review and code a seed set of documents which are then fed through the review software. By examining the reviewers’ coding of the seed set documents, the review software learns to make predictions as to the relevancy of the documents reviewed. Once the computer’s predictions and the reviewer’s coding coincide, the computer has learned enough to apply its coding to the remaining documents in the collection in an automated fashion to determine the relevancy of those documents.

Please see full alert below for more information.