 |
| This article and photos extracted from Forbes |
Ben
Kerschberg, Contributor
Fascinated by the intersection of Law & Technology.
3/23/2011 @ 10:04AM |6,348 views
E-Discovery and The Rise of
Predictive Coding
Predictive coding is the electronic coding, organization,
and prioritization of entire sets of electronically stored information (“ESI”)
according to their relation to discovery responsiveness, privilege, and
designated issues before and during the legal discovery process. Lawyers
control this process by specifying relevant criteria. Computers then expedite
discovery, as discussed in greater detail below. As I have written elsewhere,
the discovery process is becoming increasingly automated, scientific, and
objective in nature, a fact that applies both to e-Discovery with the
government (e.g., Department of Justice) and with other civil litigants.
According to Robert
Trenchard, a Partner in the New York office of Wilmer Hale, and
Craig Carpenter, Vice President and General Counsel of Recommind, an end-to-end e-Discovery and
predictive coding solution provider, predictive coding’s myriad benefits inure
to those early adopters with a risk comfort level that embraces the relative
uncertainly posed by the process vis-à-vis an attorney’s obligations to conduct
a reasonable inquiry under Rule 26 of the Federal Rules of Civil Procedure and
also to ensure attorney-client privilege in the event of inadvertent disclosure
of privileged information under Federal Rule of Evidence 502.
From Towers of Bankers’ Boxes to Predictive Coding: A Brief
History of Document Review
During the 15 months (1994-95) I spent as a legal assistant
before law school, I remember well passing conference rooms with hundreds of
bankers’ boxes, even more organized binders, tens of thousand of documents,
privilege logs that became tomes, and dozens of contract attorneys. Who was to
say at the time that the process was inefficient? It’s simply the way things
were in the age of WordPerfect. After an interim phase best characterized by
simple keyword searches and optical character recognition, e-Discovery has
evolved to predictive coding. Why has this happened? Today’s realities are far
different from what they were 15 years ago. Parties now have Terabytes of ESI.
Document review is driven by substance and relevance-centric searches that
leverage technology so as to require only a fraction of previous data
collection.
Predictive coding, according to Trenchard, is
“lawyer-driven, computer-expedited document review.” In other words, lawyers
who are intimately familiar with a case specify small sets of data that define
the crux of the case. Through an iterative search process, an algorithm
produces a much smaller set of documents. The reviewing attorney may determine
that some results are not probative and thus request that the algorithm pass
through another iteration – or many. As the computer learns to distinguish what
is relevant, each iteration produces a smaller relevant subset and a larger set
of irrelevant documents that can be used to verify the integrity of said
results by confirming the absence of any probative material therein. The extent
of the end use of that relevant set (as opposed to results by traditional
means) depends on the comfort level and risk threshold of the client and its
attorneys.
According to a highly informative webinar presented by
Carpenter and Trenchard hosted by Law.com, predictive coding is defined by at
least three defining traits. First, predictive coding leverages small samples
to find other relevant documents. Second, it reduces the amount of non-relevant
documents that attorneys must review and cull, leaving the reviewer to look at
approximately five to 20 percent of any set of documents. And third, the
results generated by predictive coding can be validated statistically.
Legal Issues Raised by Predictive Coding
Predictive coding raises at least two important legal issues
for counsel. First, can they leverage the efficiencies thereof while still
meeting their legal obligations to conduct a reasonable search for responsive
documents under the federal discovery rules? And second, how can counsel
safeguard a client’s attorney-client privilege when a privileged document slips
through the cracks and is disclosed, which happens even when the most
traditional discovery methods are used. Given the fact that predictive
coding has only come into its own in the last several years, almost no federal
(or state) case law addresses these important questions.
Federal Rule of Civil Procedure 26(g)(1)(A) requires that an
attorney responding to a discovery request attest by signature “that to the
best of the person’s knowledge, information, and belief formed after a
reasonable inquiry . . . with respect to a disclosure, it is complete and
correct as of the time it is made.” Id. (emphasis added). The salient question
then becomes what constitutes a reasonable inquiry. According to Trenchard,
several factors inform this analysis, including the importance of the issues at
stake and an examination of the proportionality of the party’s response given
the breadth of the case and the ESI being reviewed.
The second fundamental question raised by predictive coding
is whether counsel can still satisfy the mandate of Federal Rule of Evidence
502 with respect to inadvertent disclosures of material that is otherwise
protected by attorney-client privilege. With respect to “disclosure of a
communication or information covered by the attorney-client privilege or work
product protection,” id., “the disclosure does not operate as a waiver in a
Federal or State proceeding if (1) the disclosure is inadvertent; (2) the
holder of the privilege took reasonable steps to prevent disclosure,” as well
as reasonable steps to rectify the error. Fed. R. Evid. 502(b) (emphasis
added). Again, the operative standard is reasonableness, which requires that
counsel implement a document review system—irrespective of whether protective
coding is at issue—that relies on “reasonable steps,” id. at (b)(2), to prevent
disclosure. Historically, courts judging this highly subjective standard have
considered the number of documents being reviewed, the time constraints of
document production, and whether an efficient system was in place. It is
important to note that these factors guided courts in cases in which predictive
coding was not at issue.
Reasonableness and Protective Coding
According to Trenchard, it is important to remember that
perfection is not required. It is the rare party that produces every responsive
document or never inadvertently turns over a privileged document and has to
rectify the situation. Under the standard of reasonableness set forth by the
Federal Rules, the use of predictive coding is in fact reasonable. Moreover,
some of the very traits that may make it appear unreasonable at first blush are
the very factors that allow it to pass muster. For example, predictive coding
makes document review more efficient by exposing the reviewer only to those
documents that have been sorted algorithmically based on the specifications of
the producing party. In so doing, predictive coding makes it more likely that
responsive documents will be produced, thereby saving both the parties’ time
and judicial resources. Moreover the iterative nature of predictive coding
further refines relevant subsets for review, which can then be validated
statistically both for opposing counsel and the courts. There is always
considerable value to racing to the top when it comes to legal compliance. In
this case, argues Trenchard, the key is for counsel to establish mandatory
metrics that surpass any human standards that might have prevailed without
scrutiny under past e-Discovery paradigms.
___________
Ben
Kerschberg is a Founder of Consero Group LLC. Mr. Kerschberg has a Bachelor of Arts in
Foreign Affairs and German, summa cum laude and Phi Beta Kappa, from the
University of Virginia and a Juris Doctor from Yale Law School, where he was as
a Coker Fellow. He clerked for The Honorable Gilbert S. Merritt of the U.S.
Court of Appeals for the Sixth Circuit.
Ben
Kerschberg Contributor
I am a Founder of the BK Advisory Group and Consero Group
LLC. I graduated from Yale Law School and the University of Virginia and
clerked for Judge Gilbert Merritt of the U.S. Court of Appeals for the Sixth
Circuit. I contribute to Forbes, Harvard Business Review, and The Huffington
Post. My work has been featured by The Wall Street Journal's AllThingsD and
appeared in AmLaw’s Corporate Counsel, Law & Technology News, Texas Lawyer,
and Equine Journal. I'm an avid sports fan and animal lover. Please don't
hesitate to email me with comments, criticism, and ideas to
bk@bkadvisorygroup.com. Twitter = @benkerschberg. LinkedIn = benkerschberg.
The author is a Forbes contributor. The opinions expressed
are those of the writer.
Ben Kerschberg’s Popular Posts
Managing
Information Risk and Archiving Social Media 51,612 views
Makes sense that electronic discovery is on the rise, seeing as how it improves efficiency.
ReplyDelete