Despite the increasing burden of e-discovery, private litigants
and parties before the U.S. antitrust agencies have been cautious
about embracing new e-discovery technologies to assist in
identifying what documents are responsive to discovery or
government requests.The reasoning is simple: concern that the
software will miss documents that are critical to the case.This
skepticism now faces a growing body of evidence demonstrating that
the historic approach –"linear"
document-by-document review, perhaps aided by the use of keyword
searches – is no better, and likely is less accurate,
than computer-assisted review. A leading federal court has now
endorsed this empirical evidence. Last month in Da Silva Moore
v. Publicis Group, Magistrate Judge Andrew Peck of the
Southern District of New York issued an important opinion involving the use of
computer-assisted"predictive coding" in electronic
document productions.
In a variety of cases, Judge Peck has been at the forefront of
assessing challenges and solutions of e-discovery. Judicial
endorsement may spur public acceptance of these software tools by
the antitrust enforcers in the Federal Trade Commission and
Department of Justice.
Generally speaking,"predictive coding" uses human
reviewers' examination of a subset of documents
to"train" computer algorithms to review
and"predict" what other documents are responsive.
Employing this type of technology is becoming necessary to handle
the growing volume of emails and other electronically stored
information that most companies generate today. Many parties have
been wary of using new e-discovery technologies absent agency and
court approval. Relying on new technology can be especially tricky
in the context of merger review. An antitrust agency could take the
position that, having used certain technologies, the parties have
not"substantially complied" with a Second Request. This
could delay an agency decision on the merits and threaten overall
timing of the deal, and there is no suitable recourse for merging
parties.
Da Silva Moore appears to be the first court approval of
the use of such computer-assisted technology. The decision offers
some support for parties proposing to use these tools in litigation
or in response to a government agency request for documents. The
document collection in Da Silva Moore involved over three
million emails and other electronically stored information. The
defendants had proposed using predictive coding, and the parties
reached an agreement on use of the software, subject to the
plaintiffs' right to object. In approving defendants' use
of predictive coding, Judge Peck weighed (i) the parties'
agreement on the use of computer-assisted review, (ii) the large
number of electronic files to be reviewed, (iii) the demonstrated
superiority of predictive coding over traditional alternatives,
(iv) the need for cost effectiveness and proportionality called for
by the Federal Rules of Civil Procedure, and (v) the transparency
in search protocol proposed by the defendants.
Judge Peck concluded that"computer-assisted review is an
available tool and should be seriously considered for use in
large-data-volume cases where it may save the producing party (or
both parties) significant amounts of legal fees in document
review."
Although the same reasoning applies to government antitrust
investigations, as a practical matter Judge Peck's decision
will likely have more immediate influence in litigated cases:
- Private litigants are more likely to agree to use of new technologies than are the antitrust authorities. The Da Silva Moore parties had detailed parameters for the process (such as the confidence level that would be used and the size of the sample set) and engaged in an iterative process that gave plaintiffs the chance to evaluate how the software as performing. Neither the FTC nor DOJ has stated whether it would participate in similar cooperative exercises or otherwise embrace the use of predictive coding.
- Merging parties may not want to share documents with a government agency for advance vetting, as it could lead to the agency making a broader request or raise other disputes about relevance or other issues, in a context where there is no good recourse to a neutral arbiter. More than in litigation, reaching the end of a merger review usually is a priority for the parties, and this back-and-forth could consume days if not weeks.
- Without a clear statement from each agency, various staff within each agency will bring differing views on whether to accept alternative review tools, leaving parties uncertain about the outcome of a proposal to use this tool.
Nevertheless, judicial acceptance of predictive coding is a
first step along the path towards FTC and DOJ, not to mention
courts and agencies in other contexts, accepting these and other
software tools as valuable tools for e-discovery.
The FTC already has acknowledged the importance of these new
technologies. In January 2012, the FTC published proposed revisions
to the Commission's Rules (open for public comment until March
23). One area for revision is how to address electronic discovery.
Citing the widespread use of electronic materials and the need to
improve the efficiency of its investigations, the FTC
stated,"Document discovery today is markedly different than it
was only a decade ago . . . . Searches, identification, and
collection all require special skills and, if done properly, may
utilize one or more search tools such as advanced key word
searches, Boolean connectors, Bayesian logic, concept searches,
predictive coding, and other advanced analytics." The FTC also
proposed additional"meet and confer" obligations, which
could lead to transparency like that cited in Da Silva
Moore. The DOJ has not ventured into this area publicly.
The decision in Da Silva Moore has been anticipated by
lawyers dealing with large e-discovery matters, and it may spur
progress at the antitrust agencies. As Judge Peck himself
recognized, decisions like this will help clear the path for
greater use of these tools: "Counsel no longer have to worry
about being the 'first' or 'guinea pig' for
judicial acceptance of computer-assisted review."
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.