Deploying TAR and Analytics in Merger Review Proceedings
On September 21, 2022, TransPerfect Legal Solutions hosted its second annual Antitrust Clearance & Merger Enforcement Conference in Washington, DC. Industry experts discussed a range of topics including substantive developments in antitrust law and legal tech and managed review in antitrust matters.
In the presentation on Deploying TAR and Analytics in Merger Review Proceedings, moderator Michael Kriegal was joined by panelists Scott Reents, Brittany Field, and George Socha. The panelists provided an overview of technology-assisted review (TAR) methods in merger review proceedings, differences in regulatory approaches to TAR, and practical considerations for implementing a TAR workflow. The panel also discussed what we can expect for the future of TAR.
The TAR methods employed in merger review proceedings are simple active learning (TAR 1.0) and prioritized review (continuous active learning (CAL), or TAR 2.0). In TAR 1.0, a small set of documents coded by SMEs is extrapolated to the large set, while in TAR 2.0/CAL, each reviewer's codes are used to develop the constantly evolving model. The panel emphasized that both methods have their advantages, and in the case of merger review proceedings, it is often the regulator that will drive the decision regarding which method should be used.
While both the Department of Justice (DOJ) and the Federal Trade Commission (FTC) widely accept the use of TAR, there are some important differences to consider. The DOJ takes a more stringent approach to the use of TAR. It not only requires specific metrics but also prohibits the use of search terms, manual review, and email threading. Panelists agreed that a TAR 1.0 workflow is the only method that can currently accomplish what the DOJ requires. In contrast, the FTC requires written disclosure of methods used to identify responsive documents, rather than requiring specific metrics be reached. The panel acknowledged that while the FTC may move toward a more prescriptive approach, either TAR method can currently be used. The decision on which method to use for an FTC matter is best made with input from the e-discovery partner, as there are a number of factors to consider.
The decision on which TAR method is most appropriate, as well as whether to utilize additional analytics methods, such as email threading or domain analysis, depends on the amount of data, type of data, the percent of responsive documents, the deadline, and potentially other nuances of the matter. Panelists mentioned that from past experience, open communication between law firms and e-discovery partners makes for a far more successful outcome. Being open to pivoting to a more efficient workflow in light of additional information or changing circumstances also contributes to the success of the project.
Lastly, the panel considered that as we look to the future of TAR and some exciting developments, we should anticipate better solutions for dealing with mobile data and short format message data. As many companies shift away from email in favor of applications like Microsoft Teams or Slack, e-discovery technologies will need to adapt to efficiently review large volumes of message data on a compressed timeline. Challenges like these will ensure that analytics and TAR technologies are at the forefront of e-discovery innovation.
