Judge Carter’s decision in DaSilva Moore v Publicis Groupe SA
The e-discovery community has been abuzz since late last week, when District Judge Carter issued a decision in DaSilva Moore v. Publicis Groupe SA, affirming Magistrate Judge Peck’s order approving an ESI protocol that provides for the use of predictive coding. Reactions to Judge Carter’s decision ranged from euphoric (predictive coding is to discovery as jet engines were to air travel) to suspicious (the technology is overpriced, and providers claiming to have the “only” solution are not to be trusted).
My reaction falls in the middle of the spectrum. I believe Judge Carter’s order is well reasoned and reaches the right conclusion, and it provides much-needed encouragement for litigants to use automated review technologies in reducing the cost and burden of discovery. But his decision also appropriately points out the limitations of predictive coding – and all other means of reviewing documents, for that matter – and notes the safeguards litigants should put in place to bolster the reasonableness and defensibility of the processes they use.
Judge Carter observes that “[t]here simply is no review tool that guarantees perfection,” and “there are risks inherent in any method of reviewing electronic documents.” He is exactly right. Every case is unique, and requires an individualized assessment of the discovery methods potentially appropriate for the matter. Regardless of the method of document review a party ultimately chooses, they must acknowledge and understand the risks of that method and take reasonable steps to mitigate those risks. If predictive coding or other technology-assisted means of review are used, it is especially important to conduct sampling, testing and validation of the method, and prepare to present the results to the opposing party and the court if necessary.
One of Judge Carter’s reasons for approving the ESI protocol is that it “builds in levels of participation by Plaintiffs.” That’s true in DaSilva, but I don’t believe such participation by the opposing side is necessary for reasonable use of an automated review technology. Sure, cooperation in the use of technology is great if you can get it, but in the real world of high-stakes litigation, you can’t count on it. (And as a number of commentators have pointed out, involving the opposing party to the extent that they are permitted to see a sampling of documents determined by the tool to be non-responsive may not be appropriate in certain cases.) As long as a party uses automated review reasonably, defensibly, and with verifiably accurate results, they should not feel constrained by the other side’s refusal to cooperate.
Judge Carter’s cautious endorsement of predictive coding may not launch a new jetsetting age in document review. But the technology is here to stay, and hopefully his decision (and others like it) will encourage more litigants to consider using it. In many cases, it will be faster, less expensive, and more accurate than traditional methods of review.