I have a new case in federal court and I anticipate that both parties will need to sift through a high volume of electronically stored information (“ESI”) for potentially relevant documents. Opposing counsel wants the parties to take a traditional approach to the review of ESI, which involves negotiating a list of keywords and phrases related to the litigation, searching for these keywords and phrases across many thousands of electronically stored documents, and submitting the results to a large team of lawyers who will manually review each document to determine whether it is relevant.
Given the high volume of documents, I am concerned that this traditional approach to document review will yield inaccurate results and will be very expensive. I have proposed to opposing counsel that both parties use some type of technology assisted review. I am convinced a technology assisted review will be a more efficient and accurate approach. However, my “esteemed colleague” is balking. Can I compel her to use technology assisted review? On the flipside, can she prevent me from using technology assisted review?
– Curious in Cooperstown
Should you being using Technology Assisted Review?
In short, while the use of Technology Assisted Review (“TAR”) is permitted and often advisable, especially in cases with a large volume of electronically stored information, courts that have addressed the issue have held that you cannot yet force your opponent to use TAR.
However, by the same token, assuming you have a sound process in place, your “esteemed colleague” cannot prevent you from using TAR, simply because she herself chooses not to use it. In August of last year, United States Magistrate Judge Andrew J. Peck from the Southern District of New York, an acknowledged trailblazer in the world of electronic discovery, firmly held in Hyles v. New York City that a producing party in the discovery phase of litigation cannot be compelled to use TAR.
However, Judge Peck’s decision explicitly leaves the door open for a future where TAR becomes more than just a good idea and becomes mandatory. Given recent advances in technology and the acceptance of TAR by the courts, it seems entirely plausible that TAR will eventually become the most common, if not the default, way to conduct large volume document reviews.
How does TAR work?
For those readers that may be unfamiliar with the particulars of TAR, also referred to as Predictive Coding or Computer Assisted Review, it is a system that utilizes machine learning to analyze large data sets and to assist in identifying the data that is pertinent to the matter at hand. Think of it as a machine-based form of document review. TAR was developed as a tool that combines human expertise with computer learning technologies to cull the copious amount of data the average client now produces in the normal course of business (think emails, word documents, spreadsheets, calendar events, texts, PDFs, etc.). Although there are variations, the TAR process generally involves human attorneys reviewing a small set of the data at issue for responsiveness (or lack thereof) and other categories of issues specific to the case, and feeding this information to the TAR software. The computer then proliferates the coding selected by the human reviewers to the entire set of data by applying sophisticated algorithms and statistical analysis to emulate the decisions that the case attorneys would make. In this manner, by using TAR, a large set of documents can be culled down to a much smaller and more manageable set consisting of the documents that are responsive to the issues at hand.
By reducing the volume of data requiring attorney review, predictive coding offers the promise of significant cost savings where expenditures are at their highest. Take a look at these four cases that highlight the cost savings and strategic value of getting to key data faster by downloading this white paper.
While it is not a panacea or a complete replacement for human document review, under the right circumstances and used appropriately, TAR can be faster, less costly and more accurate than the traditional approach to document review. Traditionally, and particularly in the context of corporate litigation, the parties to the litigation will collect large volumes of data by applying search terms to the data sets in question and/or through interviews of the custodians of the data. The resulting data set is inevitably an over-inclusive set of documents that then need to be reviewed for relevance, privilege, confidentiality, etc., typically by enlisting large document review teams to manually inspect and code each and every document in the collected data set. In the context of large corporate matters, the documents to review can literally number in the millions.
Judge Peck’s Stance on Using TAR
In 2012, in what he later described as his “seminal” TAR case, Da Silva Moore v. Publicis Groupe, Three years later, in Rio Tinto PLC v. Vale S.A., Judge Peck reaffirmed his approval of TAR, noting that “the caselaw 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.” Judge Peck also used Rio Tinto as an opportunity to de-mystify TAR and cast it as just another tool to manage electronic discovery that “[should not be held] to a higher standard than keywords or manual review.”
However, until Hyles v. New York City, it remained unclear whether a producing party who preferred to use keyword searching could instead be compelled to use TAR over its objections. Judge Peck has answered this question with “a decisive ‘NO.’” In Hyles, the plaintiff, a New York City employee who asserted various discrimination and hostile work environment claims, sought to compel her former employer, the City of New York, to use TAR. New York objected, preferring instead to generate responsive documents using keyword searching, in part, perhaps, because it had a large, salaried staff that it could devote to the large document review project.
Judge Peck agreed with the plaintiff that TAR is the best and most efficient search tool for most cases and that a production using keywords rather than TAR “may not be as complete.” Judge Peck even took it a step further and conceded that the Court itself “would have liked the City to use TAR” because, in Judge Peck’s opinion, it is an effective and efficient tool. Nevertheless, Judge Peck held that the Court could not force New York City to use TAR against its will. Judge Peck noted that this did not leave the plaintiff without remedies: if the plaintiff could later demonstrate deficiencies in New York City’s search term generated document production, New York City might be required to re-do its search.
Interestingly, Judge Peck noted in his closing paragraph of Hyles that “[t]here may come a time when TAR is so widely used that it might be unreasonable for a party to decline to use TAR. We are not there yet.” Whether and when we arrive “there” remains to be seen. But for now, it seems safe to say that TAR is here to stay and is likely to become more widely used in the future.
I agree with Judge Peck and, while you may not be able to compel your opponent to use TAR at this time, you should always consider whether it is appropriate to your case: a forward-thinking approach will serve you and your clients well as this technology gains broader acceptance. It is important to note that, if you do in fact choose to use TAR for your own case, a sound process is necessary to make sure your implementation and resulting document production is conducted in a defensible manner. This often involves a combination of TAR, manual review, and the structuring of quality control workflows to ensure that the scope of review is comprehensive and the results are validated. If you want to use TAR, and structure the process in the right way, Judge Peck’s decision sends a strong message to your opponent that she would have a difficult time preventing you from doing so, even if she herself chooses not to use it. Good luck!
Attract & Retain Top Talent
With a rapidly changing industry, it's vital to offer the right compensation and set the right expectation. With our Salary Guide, get detailed job descriptions, industry insights and local salary data to equip your managers with hiring confidence and expertise.Get your copy »