Since the first judicial opinion endorsing the use of Technology Assisted Review (or TAR) was written by Judge Andrew J. Peck in 2012, an entire legal industry has grown up on the premise of streamlining the document review process in discovery – that is, taking a repetitive task traditionally performed entirely by attorneys and introducing the concept of computer assistance to increase efficiency and improve consistency. And while some of the marketing efforts surrounding TAR make it seem TAR can replace attorney review wholesale, some attorneys considering the use of TAR in litigation or in response to regulatory inquiries are more focused on ignoring the “buzz” and instead have zeroed in on what judges have actually said (and held) regarding TAR.
The recent article “Court Guideposts for the Path to Technology Assisted Review Adoption,” published in Computer Science and Information Technology and originally presented as part of the proceedings of the 2017 Georgetown Advanced eDiscovery Institute, addressed exactly that issue. In particular, the article considered four primary concepts that judges either opined on or addressed in detail when issues of TAR came up during matters:
- Proportionality and Party Objections
- Cooperation, Transparency, and the Provision of Seed Sets
- Recall and Precision
- Audit Practices or Validation
Examining those issues, the article determined where the marketing hype stopped and where specific court direction began.
For proportionality, courts enthusiastically allowed the use of TAR, while stopping short of requiring it. But while courts would not force parties to utilize TAR, at least one court noted that TAR was a proven option that “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.” And just as courts would not prescribe the use of TAR, neither would courts hold that a party could not use it, even in the face of party objections. Courts addressing this subject balanced the issue carefully, essentially holding and recognizing that TAR did promote efficiency while simultaneously confirming that, at least at this point, parties would not be required to do so.
In service of party cooperation, courts examined the issue of transparency – which, in the context of TAR, also may include the provision of seed sets or those documents certain types of TAR processes use to “train” the review system. Here, too, courts lauded those instances where parties decided affirmatively to share this information in service of efficiency but rarely required such openness. Aside from that exception, most courts addressing the issue then indicated strong support for those parties who shared seed set documents, information regarding the process or sufficient information so that the parties understood clearly how TAR was used in a given matter.
The concepts of recall and precision measure the effectiveness of the TAR process as well as gauge the sufficiency of a party’s production after TAR use. Following a clear trend within party practices, when parties raised issues associated with recall and precision measures, the courts allowed parties to determine their own direction, noting only that the metrics the parties choose should be reasonable and stating firmly that perfection is not the standard by which party efforts will be judged. Finally, courts examined those audit practices producing parties incorporated into their proposed TAR use or those receiving parties raised after the TAR process was complete. Overall, courts allowed parties to develop reasonable audit practices, and then they tacitly endorsed the use of those practices in support of an admittedly “imperfect” process.
As noted in the article and echoed here, for “those parties considering the implementation of TAR as a discovery mechanism, the published case law does provide guidance for a sound protocol and approach, and indicates how courts will consider the specifics of a party’s ultimate determinations and resulting productions.” Ultimately, courts considering the issue of TAR gave direction to parties that focused not on party wrongdoing or mistakes but instead on the correct way to use TAR in the service of the matter and of discovery generally.