Computer Assisted Review in Discovery; Court of Chancery First Court to Order It
In another Delaware “first in the nation,” Vice Chancellor Laster of the Court of Chancery on October 15, 2012, in EORHB, Inc. v. HOA Holdings LLC (C.A. No. 7409-VCL) ordered the parties to “show cause” why computer assisted review should not be used for discovery of electronically stored information (“ESI”) in that matter. After a hearing on a motion for partial summary judgment and a motion to dismiss a counterclaim, Vice Chancellor Laster, sua sponte, raised the issue of computer assisted review in discovery for the balance of the case, saying: “[t]his seems to me to be an ideal non-expedited case in which the parties would benefit from using predictive coding. I would like you all, if you do not want to use predictive coding, to show cause why this is not a case where predictive coding is the way to go.” Transcript at 66.
Background on Computer Assisted Review
Computer assisted review (“CAR”) (also known as predictive coding or technology assisted review) is a new approach to the traditional linear document review which can save the parties significant amounts of time and money while at the same time greatly increasing the likelihood of accessing relevant information. CAR, which is a combination of human review and artificial intelligence, involves the implementation of software tools that use sophisticated algorithms (sets of rules that define a sequence of operations) to enable the computer to determine the scope of relevance based on training by a human reviewer. Computer assisted review involves coding by a small team of trained reviewers who review and code a sample set or “seed” set of documents ( a thousand to two thousand documents is a normal amount). That coding information and the data to be reviewed is loaded into a database and using a software program, the computer identifies common properties of those documents. As the human reviewers continue to code more sample documents and load both the documents and the coding information into the database, the computer will start to predict the reviewer’s coding for relevance determination. When the computer’s predictions and the reviewer’s coding coincide at a sufficiently high pre-determined level, the system will be deemed to have been trained to make confident predictions about the relevance of the remaining documents.
CAR is a very “hot topic” in e-discovery and it started in February 2012 with Federal Magistrate Judge Peck in Monique Da Silva v. Publicis Groupe & MSL Group, Case No. 11-cv-01279 (S.D.N.Y. Feb. 24, 2012). In DaSilva, at the request of the parties, the court approved the use of predictive coding to locate relevant ESI. Da Silva is remarkable because it was the first written decision where a judge approved any protocol dealing with CAR. It is also important to note that Judge Peck did not initiate the discussion about CAR in that case. Rather, the parties brought the topic to the judge telling him that they had agreed to the defendants’ use of it but disagreed as to the scope and implementations, so the court was asked to decide the issues related to the protocol. In EORHB, Vice Chancellor Laster raised the issue of CAR on his own and ordered the parties to use CAR or show cause why that technology should not be used.
Computer assisted review is most helpful in situations where the lawyers understand the process, the stakes are large, and the volume of data is significant – like Da Silva, where over three million documents were involved. Small-stakes cases will not benefit from this technology so they will have to keep using a linear/manual approach to review and search terms. Unfortunately, studies have shown that manual review is much less reliable than computers and can miss a surprising amount of relevant and privileged information even during the most diligent document review.
While this is not the first time that Vice Chancellor Laster, who is a tech-savvy jurist, has suggested computer assisted review to parties in matters before him in complex corporate matters, this is the first time that any court has required the parties to show cause why they would not benefit from using computer assisted review. While the judge used the phrase “predictive coding” a number of times in the transcript, in context he was referring to technology or computer assisted review and not a particular vendor. He went on to tell the parties that they should select an appropriate vendor to handle the computer assisted review process and if they could not agree on a single vendor to submit names to the court and the court would decide. It is also important to point out that the parties can come back to the court and explain why computer assisted review in this case would not be practicable or feasible. This is certainly a clear indication that the use of computer assisted review is on the minds of many judges who are looking for ways to bring the problems of ESI in line with the mandates of Rule 1. And as we see here, sometimes the parties and their lawyers need to be strongly encouraged to consider the use of technology as a means to reduce costs and increase efficiency in document review involving electronically stored information .
Editorial Comment and “Hat-Tip” to Vice Chancellor Laster
Vice Chancellor Laster has created quite a buzz in both the e-discovery world and the corporate litigation world by ordering CAR, but for very different reasons. The corporate litigation world is buzzing because CAR is not a topic that is on their radar for the most part. The e-discovery world is very surprised to hear that a judge has ordered CAR. It is true that prior to Vice Chancellor Laster’s instructions, no judge in the country had ordered the parties to use CAR. However, in my opinion, it is a great for the advancement of the use of technology in litigation for a judge on the Delaware Court of Chancery to do it because companies and their counsel (in-house and outside) have been slow to understand let alone embrace the use of technology in litigation. The Federal Rules of Civil Procedure were amended six years ago to address the problems associated with ESI and even conservative estimates puts the percentage of lawyers who are not well-versed in e-discovery to be over 50%. It is going to take a proactive bench (Federal and state) to get the word out and create incentives for lawyers to embrace these concepts as well as the technology. So a “hat tip” to Vice Chancellor Laster for getting the bits and bytes rolling and making the topic of computer assisted review relevant in Delaware and beyond.