It has been three years since Magistrate Judge Andrew Peck issued his February 2012 decision in Da Silva Moore v. Publicis Groupe & MSL Group, 287 F.R.D. 182 (S.D.N.Y. 2012), judicially accepting for the first time the use of technology-assisted review (“TAR”). In Da Silva Moore, Judge Peck recognized that TAR, also known as computer-assisted review or predictive coding, is an acceptable way to search for relevant electronically stored information. He has recently issued an opinion in Rio Tinto PLC v. Vale S.A., 14 Civ. 3042, 2015 WL 872294 (S.D.N.Y. March 2, 2015), titled “Da Silva Moore Revisited,” which addresses how TAR has been treated by courts since Da Silva Moore, and whether parties must disclose seed sets used to train the TAR tool.
In analyzing the current judicial landscape in Rio Tinto, Judge Peck states that “the case law has developed to the point that it is now black letter law that where the producing party wants to utilize TAR for document review, courts will permit it.” Indeed, he notes Sedona Principle 6’s guidance: “Responding parties are best situated to evaluate the procedures, methodologies, and technologies appropriate for preserving and producing their own electronically stored information.”
However, Judge Peck also highlights that one TAR issue remains open: how cooperative and transparent must the parties be with respect to the seed sets used to train the TAR tool? Judge Peck mentions some of the cases in which parties agreed to produce their seed sets. Rio Tinto, 2015 WL 872294 at *2, citing Bridgestone Ams., Inc. v. IBM Corp., 2014 WL 4923014 (M.D. Tenn. July 22, 2014) (plaintiff offered to provide seed set); In re Actos (Pioglitazone) Prods. Liab. Litig., 2012 WL 7861249 (W.D. La. July 27, 2012) (parties’ protocol had “experts” from each side coding the seed set); Da Silva Moore, 287 F.R.D. at 187 (defendant volunteered transparency and represented that all documents reviewed for the seed set would be turned over, aside from privileged documents). He also referenced a bench decision in Federal Housing Finance Agency v. HSBC North America Holdings Inc., Case Nos. 11 Civ. 6189-6190, 6193, 6195, 6198, 6200-6203, 6739, & 7010 (S.D.N.Y. July 24, 2012), in which the court explicitly required transparency and required that plaintiff have access to the seed set’s responsive and non-responsive documents, except privileged documents.
Conversely, Judge Peck recognizes that other courts have found no authority permitting courts to require a party to share its seed set. Rio Tinto, 2015 WL 872294 at *2, citing In re Biomet M2a Magnum Hip Implant Prods. Liab. Litig., 2013 WL 6405156 (N.D. Ind. Aug 21, 2013). In addition, Judge Peck cites to a recent law review article authored by The Honorable John M. Facciola and Philip J. Favro addressing the debate regarding whether seed sets should be produced. See “Safeguarding the Seed Set: Why Seed Set Documents May Be Entitled To Work Product Protection,” 8 Fed. Cts. L. Rev. 1 (2015).
Importantly, Judge Peck proceeds to instruct that, even without production of seed sets, requesting parties can employ other means to help ensure that TAR training and review were conducted appropriately. He suggests methods to assess and evaluate TAR, such as statistical estimation of recall at the end of the review, determining whether there are gaps in the production, and quality control review of samples from the set of non-responsive documents.
In the end, Judge Peck approved the parties’ TAR protocol in Rio Tinto, in which they agreed to disclose all non-privileged documents in the seed sets. Because the parties agreed to the protocol prior to submitting it to Judge Peck, the court did not set forth a position on whether it would order production of a seed set absent agreement.
While revisiting Da Silva Moore, Judge Peck also provided general advice regarding TAR: “One point must be stressed – it is inappropriate to hold TAR to a higher standard than keywords or manual review. Doing so discourages parties from using TAR for fear of spending more in motion practice than the savings from using TAR for review.” While Judge Peck’s opinion does not expressly address whether parties need to obtain opposing parties’ consent or court approval before using TAR, this language would seem to imply that no such consent or approval is a prerequisite, since there is no similar approval requirement for the traditional use of keywords and review methodologies. Indeed, courts generally leave it to the parties to decide how best to respond to discovery requests. See Rio Tinto, 2015 WL 872294 at *1, citing Dynamo Holdings Ltd. P’ship v. Comm’r of Internal Revenue, 2014 WL 4636526, *3 (T.C. Sept. 17, 2014) (stating that “the Court is not normally in the business of dictating to parties the process that they should use when responding to discovery”).
Consequently, we are left with one more opinion approving the use of TAR; but, the debate will continue as to how much parties should “meet and confer” before employing such technology, and whether courts have the authority to require production of seed sets without party agreement. How these issues play out and how much additional motions practice will result from increasing use of TAR for litigation discovery remains to be seen.