Washington, DC – With the exponential increase in the volume of Electronically Stored Information (ESI) and the emergence of cloud technology, the mushroom clouds of litigation have thickened over the industry, overshadowing the process of seeking, locating, securing and searching of its electronic data with the intent of using it as evidence in a civil or criminal legal case.
The sluggish nature of human reviews, combined with skyrocketing price tags and the intricacies of systems in other countries, have resulted in a much needed emerging method of TAR (Technology-Assisted Review), which is being recommended by experts.
India America Today decided to get a first hand perspective and in the second of a series of articles exploring the subject of e-discovery, Tejinder Singh, Editor of India America Today, spoke to Jason R. Baron, Director of Litigation, National Archives and Records Administration (NARA), in College Park, Maryland, on the ongoing debate about the use of TAR.
What is your perspective on the new technologies like TAR (Technology Assisted Review), which is being recommended to deal with the review process of e-discovery?
Let me first thank you for the opportunity to speak. I need to state right at the top that any remarks I say here represent my own personal views, and not necessarily the official position of either my agency or the US Government.
As you know, I have been a strong advocate for the legal profession, including lawyers inside government as well as in the private sector, adopting various forms of more advanced search methods for use in e-discovery.
Over the past decade or so, I have given hundreds of talks, including appearances at e-discovery conferences and international workshops, with an aim towards educating the profession on the limitations of keyword searching, while at the same time urging consideration of more advanced search methods.
Today, at least some of these methods go under the label “Technology Assisted Review.” In my opinion, new methods and smarter thinking are necessary to reduce needless burden and expense involved in conducting e-discovery searches for ESI.,
Then why are there no signs of implementation by government agencies?
I’ve seen some progress, but as you know that the government has limited resources, and we are in a climate where agencies face tremendous budget pressures. To the extent that TAR methods are expensive to implement,it cannot be reasonably expected that many or most government agencies will be able to accomplish doing so in the short term.
Are you keeping abreast of ongoing developments in TAR and other programs?
I am, both to assist my own agency, as well as a basic matter of staying up to speed on latest developments that any lawyer involved in the practice of litigation and e-discovery needs to know about. More specifically, my agency recently was involved in a solicitation to industry to inform us of latest developments in search technologies, a development I applaud.
When did this happen and are you still entertaining vendors to come and show their wares?
NARA put out an RFI (Request For Information) in Fed Biz Opps.gov in March 2012, calling for “Advanced Search Technology Solutions Capability Statements.” We followed up giving opportunities to present about existing and emerging technological solutions that could support potential advanced search requirements (e.g., conducting e-discovery or enterprise search) of archival records in NARA’s legal custody, as well as the operational records of the agency. I anticipate that there may be additional opportunities in the future to present, including through NARA’s TechWatch presentations forum.
Speaking more generally, I believe there is a growing interest in exploring new methods such as TAR both in government and the private sector.
Do you have any TAR systems already in place?
We do not employ TAR methods as such at the National Archives, as we have not yet implemented the software to do that. Other places in government I believe exist that have at least experimented with what is known as “predictive coding.” I don’t know the specifics, as it’s all anecdotal.
You mentioned the budget, but according to experts, TAR is aimed at reducing the cost of review. So does it have a future?
Yes, it has a future. It has been shown repeatedly to reduce document review cost by using computers to code vast quantities of documents which otherwise have to be coded by people.
However, there is no free lunch in this world and one needs to purchase the software, implement it, and continually maintain it over time. Those are not inconsiderable costs, including front end costs, even if you are saving money in a particular litigation.
For a government agency to buy lots of software and have it sit around waiting for some future use, is something that needs to be evaluated based on the degree of litigation exposure a given agency faces.
So, I am a big advocate of it, but I realize that every institution that is considering using the software and evaluating the market needs to do a cost-benefit analysis.
What areas of EDRM process need to be focused for e-discovery to make litigation support less expensive and more effective?
It is well known that document review, the processing portion of the e-discovery, is where the giant costs are, both to determine relevance and privilege. Any method that reduces the amount of expenditure, time, resources, and people on document review is a good thing. But every stage of the process could benefit from more advanced methods.
What’s next on the horizon in this area?
My crystal ball is a bit cloudy, but there is no question that technological progress in the field of artificial intelligence has the potential to continue to revolutionize the discovery process. Over the past several years, I have had the privilege of a hosting a series of international workshops bringing together members of the AI, information retrieval, and legal communities to discuss advanced search topics.
The next workshop, labeled DESI V (for Discovery of ESI), is scheduled to be held next year in conjunction with the AI conference known as ICAIL 2013, in Rome, Italy, in June. I invite interested readers of your column to look for a call for papers later this year, as the workshop is intended to focus on standards for predictive coding and supervised learning technologies. I’m also happy to correspond with interested individuals, and can be reached at [email protected].