The latest e-discovery opinion to cause a stir is from a federalmagistrate judge well known in the field for his ability tocommunicate e-discovery “best practices” to practitioners. InDa Silva Moore, et al. v. Publicis Groupe & MSLGroup,[1] Magistrate Judge Andrew J. Peck issued the firstreported opinion dealing with a technology called “predictivecoding.”

While Judge Peck's comments at the hearing preceding the writtenopinion were quickly heralded by vendors as officially “validating”predictive coding as a judicially endorsed software product, theopinion specifically rejects this misinterpretation.[2] Rather,Judge Peck states that predictive coding is an appropriate tool forcertain cases but should be subject to the same kind of rigorousanalysis and testing as other methods of document review:

[I]f the use of predictive coding is challenged in a case beforeme, I will want to know what was done and why that produceddefensible results. I may be less interested in the science behindthe “black box” of the vendor's software than in whether itproduced responsive documents with reasonably high recall and highprecision.

The underlying case involves claims of gender-based employmentdiscrimination. In the course of discovery, the parties disagreedon the appropriate e-discovery protocol—not with respect to the useof any particular technology, but about such mundane matters as howmany documents should be reviewed, which custodians' email shouldbe searched, which custodians' documents should be reviewed when,discovery cutoff dates and what sources of electronic informationshould be searched.

Judge Peck's opinion advises attorneys that predictive codingcan be an acceptable aid in conducting document review inappropriate cases, as long as it is part of a process that isdefensible, and subjected to quality control testing appropriatefor any document review, however conducted. These kinds of qualitycontrol methods have previously been elaborated by the SedonaConference, in its white paper entitled “Commentary on AchievingQuality in the E-Discovery Process,” and in the New York State BarAssociation's “Best Practices in e-discovery in New York State andFederal Courts.”

Predictive coding is a term that has become a buzzword ine-discovery, but seems to have various meanings depending on whichsoftware vendor is speaking. Vendors often use the term inconjunction with the enticing moniker “automated review.” JudgePeck uses the term “computer-assisted coding” and defines it as“tools … that use sophisticated algorithms to enable the computerto determine relevance, based on interaction with (i.e., trainingby) a human reviewer.”

In Da Silva Moore, Judge Peck asked the defendants toagree on a transparent protocol that would include producing thedocuments used to “train” the software, i.e., the seed documents,as well as the other parameters of the process. While Judge Peckdid accept predictive coding as an acceptable tool in this case tofacilitate e-discovery, the underlying condition of acceptance—thatthis acceptance would be contingent on whether it was defensible bybeing “quality control verified” —would be advisable in anydocument review technology or process:

As with keywords or any other technological solution toe-discovery, counsel must design an appropriate process, includinguse of available technology, with appropriate quality controltesting, to review and produce relevant ESI …

The underlying reasoning is as clear as it is incontrovertible:if you can verify that the results are accurate, which technologygenerated them is a moot point.

The bottom line here is that there is more than one way to skina cat. The trick is to verify with a high degree of confidence thatthe cat has in fact been skinned. The application of any particularsoftware in conducting document review has to be part of a processin which a number of elements apart from the software are used toarrive at the right results. The arrival of the e-discovery “easybutton” has been delayed. Human judgment is still part of theprocess, and we all know that reasonable minds willdiffer—especially in litigation.

The views expressed herein are those of the authors and donot necessarily reflect the views of Ernst & YoungLLP.

[1] 1:11-cv-01279-ALC-AJP (S.D.N.Y. Feb. 24, 2012).

[2] Judge Peck noted that the vendor of the software used bydefendants had initiated a press release mischaracterizing theopinion as ordering the parties to use predictive coding. Theparties had already agreed to its use, and Judge Peck accepted thisagreement.