Facts
- Monique Da Silva Moore and other female employees sued Publicis Groupe and its U.S. public-relations subsidiary, MSL Group, alleging company-wide gender discrimination in pay, promotion, and employment actions under Title VII and related federal, state, and local laws, including the Equal Pay Act, FLSA, and New York statutes.
- The case involved a very large volume of potentially responsive electronically stored information (ESI), reported to exceed three million items, largely email.
- Pretrial matters, including discovery, were referred to U.S. Magistrate Judge Andrew J. Peck.
- The parties needed multiple court conferences to work out an ESI search and review protocol that could yield high statistical confidence in the results.
- MSL proposed using technology-assisted review (TAR), specifically predictive coding (computer-assisted review), to identify responsive documents for production.
- MSL’s counsel and technical personnel disclosed details about the planned TAR approach during the meet-and-confer process and conferences, which helped the parties finalize an agreed process.
- The parties entered into a stipulation setting out the predictive-coding protocol; the court so-ordered it. The stipulation noted that plaintiffs objected to the protocol but signed it as reflecting the court’s rulings on disputed points.
- Plaintiffs objected to using predictive coding and to elements of the protocol, arguing it was not sufficiently tested or reliable and seeking changes to the process.
Issues
- Whether computer-assisted review (predictive coding) is an acceptable method for searching and reviewing large quantities of ESI for responsiveness under the Federal Rules of Civil Procedure.
- Whether, given the size of the ESI collection and the negotiated quality-control measures, the predictive-coding protocol ordered by the court was a reasonable discovery method in this case.
Decision
- The court held that computer-assisted review, including predictive coding, is an acceptable method to search for relevant ESI in appropriate cases.
- The court approved the parties’ predictive-coding protocol for MSL’s ESI review and production, despite plaintiffs’ objections.
- The court treated the order as a discovery-management ruling in an ongoing case, setting the method for review rather than resolving the merits.
Legal Principles
- The Federal Rules require a reasonable search for responsive information, evaluated in light of proportionality and practical limits, not a guarantee of perfect recall.
- Rule 1’s directive for the “just, speedy, and inexpensive” determination of actions supports using methods that can reduce cost and time in large ESI reviews.
- Computer-assisted review is not disfavored simply because it uses algorithms; it may be used when paired with sound quality controls such as iterative training, sampling, and validation.
- Cooperation and disclosure about search methods and testing can reduce disputes and support judicial approval of an ESI review protocol.
- Courts have broad discretion to manage discovery and to approve negotiated protocols that are reasonably designed to locate responsive ESI while controlling burden and expense.
Conclusion
In Da Silva Moore v. Publicis Groupe, the S.D.N.Y. approved a negotiated predictive-coding protocol for reviewing and producing a multi-million-document ESI collection, holding that computer-assisted review can satisfy the Federal Rules when the process is reasonably designed, tested through quality-control measures, and proportionate to the discovery burdens in the case.