Our latest post about HSR Second Requests discussed five key considerations for HSR Second Requests. We promised an additional “bonus” consideration in our next post and that consideration is whether to consider using Technology Assisted Review (TAR) to assist with your Second Request review.
The use of TAR for eDiscovery review has become a standard and accepted approach within the legal community and has been court-approved for over ten years as an acceptable approach for litigation document review. There are benefits to using TAR in HSR Second Request reviews, but there can also be challenges. Let’s discuss each in turn.
Benefits to Using TAR in HSR Second Requests
There are at least three benefits to using TAR in HSR Second Request reviews. They include:
Speed: With HSR Second Requests having a much more compressed timeframe for response (compared to typical litigation timeframes), there is a need for speed in conducting the review. The machine learning aspect of TAR enables review to be conducted more quickly because it typically requires human review of only a fraction of the data set to train the algorithm to make responsiveness determinations. With the tight timelines of a Second Request production, using TAR to either decide or suggest which documents are responsive to the Second Request can be a game changer.
Reduced Cost: Perhaps the primary reason that organizations use TAR to conduct document reviews is the cost savings. Not only does limiting the human review to algorithm training save time, but it also saves considerable costs for most reviews. Second Requests typically call to produce an unprecedent volume of data. Even though an HSR Second Request is a high-stakes, “make-or-break” activity for your merger or acquisition, it still makes sense to manage the costs of review if you can do so without sacrificing quality.
Accuracy: Speaking of not sacrificing quality, it has become generally accepted that TAR is also more accurate in most cases than manual reviews in large-scale document collections involving a team of reviewers. While attorney reviewers do their best to accurately review documents, several studies over the years have illustrated that documents are often misclassified in human review. Human attention waivers and each human may have a different interpretation of responsiveness (despite efforts to standardize that understanding). Assuming proper training of the machine learning algorithm, TAR is generally more accurate than human review. The algorithm never gets distracted!
Challenges to Using TAR in HSR Second Requests
There are also at least three challenges to consider when using TAR in HSR Second Request reviews. These challenges can be overcome by an experienced service provider partner in using a hybrid approach to review various data types. They include:
Limited Review of Certain Document Types: While TAR is great for classifying text-oriented documents, it’s typically not good with spreadsheets and other documents that are number-intensive, as opposed to text-intensive. With a lot of HSR Second Requests involving scrutiny of financial documents, TAR may be less useful for a portion of the collection. Of course, TAR is also no help with files that contain no text, such as images, audio and video files. Even if you use TAR, you may be forced to conduct a side-review of file types that TAR is unable to classify.
Nuanced Decisions More Difficult: While TAR eliminates mistakes made in clear cut responsiveness determinations (which comprise 90-95% of determinations), it can struggle with those determinations that are more nuanced. This is the one area where human review can have an advantage over TAR. If there are a lot of nuanced responsiveness decisions to be made in your Second Request review, TAR may need to be supplemented with some experienced human reviewers.
Disclosure: As is the case with TAR in litigation reviews, you may need to disclose the use of TAR and provide additional information about the review process, especially for the DOJ. It makes sense to understand the requirements of the regulator when using a TAR review and inviting your service provider partner to the discussions to establish an agreed upon protocol.
So, TAR or no TAR? We’ll give you the lawyerly answer of “it depends”. There are several factors that can impact your TAR decision, including timeframe, information being requested, composition of your document collection and more.
It’s best to work with an expert that understands the HSR Second Request process, the capabilities of TAR and how to manage TAR workflows, as well as how to apply the parameters of your situation to make the decision about when and how to use TAR. The decision whether to use TAR or not should be made without expert advice – the stakes are too high to make a poor choice for your Second Request!
For more regarding Cimplifi corporate transactions & antitrust capabilities, click here.