TAR and The Case Law
TAR is part of the fabric of electronic discovery. It represents a natural part of the movement toward making discovery possible without cataclysmic economic consequences and the toppling of a legal system based on the search for the truth. Without TAR, the scope of legal discovery would either have to narrow dramatically or parties would find themselves unable to comply with discovery obligations for lack of money and time. Even with TAR, fundamental views of discovery have been forever altered as a result of ever-increasing data volume, variety and velocity, or "Big Data."1 Principles such as proportionality, transparency, and even cooperation are widely (although not universally) viewed as essential if electronic discovery is to continue to be part of our legal system. TAR is well suited to adhere to these principles.
There are published opinions where the use of TAR has been the subject of discovery disputes, and there are cases where courts have "recommended" to parties that they use TAR to relieve burdensome electronic discovery obligations. However, there is still little in the way of published law directly addressing TAR (and nothing in the way of authoritative precedent), despite the bellwether, high drama of Moore v. Publicis Groupe, et. al.,2 discussed below. Nevertheless, it is possible to make educated guesses about where the trend is heading in terms of court attitudes toward TAR, not only through interpretation of case law but also through thought leadership publications and dialogue at conferences and in other continuing legal education settings. The trend is undeniably toward promoting greater use.
As a sign of the times, consider the comment submitted by the Duke University School of Law Center for Judicial Studies, dated October 17, 2013. In response to the proposed amendments to the Federal Rules of Civil Procedure published for public comment on August 15, 2013, by the Civil Rules Advisory Committee, it proposed that the following sentence be added to the committee note to Rule 26(b)(1):
As part of the proportionality considerations, parties are encouraged, in appropriate cases, to consider the use of advanced analytical software applications and other technologies that can screen for relevant and privileged documents in ways that are at least as accurate as manual review, at far less cost.
One of the great debates in electronic discovery with particular relevance to TAR is about the extent to which producing parties should be required to disclose the methods, as well as technologies, that underpin their productions. Differing approaches to such disclosure are taken to some degree by Magistrate Judge Peck in Moore and Magistrate Judge Nolan in Kleen Products.3 On one side of this debate are those who argue that this kind of transparency has never been required in the past, so there is no reason the use of TAR should change the existing paradigm. They point out that, historically, requesting parties were not routinely allowed to inquire into a producing party's document selection methods or criteria unless they could show that there was some kind of defect in the production. Even then, the inquiry would be limited to the particular defects shown, as opposed to a general inquiry into how documents were identified for production. It is arguable that a certain degree of opacity is justified because inquiring past a certain point would reveal the attorney's thought process, which is a classic "work product" that is protected from production.
The opposing view holds that electronic discovery has changed these dynamics. Given the amount of filtering that necessarily takes place between an original universe of documents and what is ultimately produced, as well as the complexities involved in performing all manner of legally necessary operations on ESI, from preservation through production, huge gaps in production could arise from small decisions, computational errors, misconfigurations of technology, search terms, etc. Clearly, TAR adds additional opportunities for divergence and disagreement. Moreover, holders of this view argue that they just want to know what was done-a factual matter-as opposed to why it was done.
There is little question that opponents of transparency appear to have an uphill struggle. This was clear with the amendments to the Federal Rules of Civil Procedure enacted in 2006. Rule 26(f) was modified to require that the "meet and confer" include a discussion of electronic discovery issues. While there was no specific direction as to what (or even whether) parties need to discuss how they arrived at their document production, the clear mandate is that greater cooperation and transparency is a necessity if electronic discovery is not going to hopelessly bog down an important element of our legal system. The proper application of this mandate to TAR, however, is the subject of debate.
The idea behind requiring more up-front discussion about electronic discovery is to prevent disagreements from derailing the process and delaying adjudication of the issues. Given the effort and expenditure involved in obtaining and migrating ESI from preservation through production, finding out after the fact that the entire effort will need to be done differently would be, to put it mildly, inefficient. As demonstrated in some of the cases discussed below, the idea that a party should redo a major production effort because there was purportedly a more accurate way of doing things, or because there was a flaw in the specific TAR process, raises the hackles of those involved in the process-or even more, of those who pay for it. Rule 26(f) shows that courts expect parties to discuss what they need to in advance to avoid situations where many miles down the litigation road, and millions of client dollars later, they discover there was a wrong turn and the steps must be retraced.
Notwithstanding the amendment, a constant refrain has been heard from the judiciary that parties are not adhering to the new rule and instead are still conducting "drive-by" 26(f) conferrals. In other words, lawyers are not adequately addressing the electronic discovery issues that could grind the case to a halt later. An increasing number of opinions chastise parties for failing to raise electronic discovery issues at the meet and confer stage and refuse to grant relief because of this failure.4 The upshot is that an ill-considered decision to avoid transparency early in a matter comes at the risk of eventually getting stuck somewhere unpleasant (and expensive) to remedy the problems that result.
The electronic discovery processes used by parties are too frequently subject to discovery themselves to insist that the old ways of doing things should persist. The train has left the station. Electronic discovery processes should not be something you are ashamed of or afraid to show your adversary. Chances are you will have to do so sooner or later, either as attorney work product or in a theoretical debate over the most appropriate approach.
With respect to TAR in the broad sense, many opinions in recent years show a judicial expectation of open discussion of and attempt to find agreement on search terms.5 At a minimum, the notion that electronic keyword search terms used to find responsive documents should be protected as work product may be fading fast. Generally, this level of disclosure of the document selection and review process has become a routine and expected part of electronic discovery procedure in many quarters, although opponents of such transparency are still mounting vigorous defenses when it comes to specific aspects of TAR, for example, "seed sets."
And then there is the Sedona Conference's "Cooperation Proclamation."6 This document, weighty by virtue of the many judicial signatures it has attracted, mandates that all lawyers should just try to get along when it comes to electronic discovery. Lawyers who resist sharing their electronic discovery processes risk being branded uncooperative, and no prudent lawyer wants a judge to think they are uncooperative. Cases described below (as well as others) have cited this proclamation, and any debate about whether cooperation is consistent with the lawyer's duty to be a zealous, vigorous, or otherwise enthusiastic advocate for the client is merely academic. The finer distinctions about what level of disclosures in the spirit of cooperation are consistent with the adversarial system can be argued, but the appearance of noncooperation will not help anyone get the TAR solution they want.
2.2 Emerging Case Law
2.2.1 Judicial Recognition of TAR-Moore7
While discussion of TAR in the electronic discovery "community" started in earnest long before (actually, many years before) it appeared in any judicial opinion, the proverbial stake in the ground was driven by a Magistrate Judge in federal court in New York. This judge was known for his active involvement in electronic discovery, through his judicial opinions and through participation in thought leadership conferences and publications (a factor that became a subject of contention in the case discussed below). Moore involved gender and pregnancy discrimination claims against large advertising companies. In Moore, Magistrate Judge Andrew Peck of the U.S. District Court for the Southern District of New York unambiguously held that there is no a priori reason that the use of TAR...