Technology Assisted Review (TAR), also known as predictive coding or computer-assisted review, has been defined as “[a] process for prioritizing or coding a collection of documents using a computerized system that harnesses human judgments of one or more subject matter expert(s) on a smaller set of documents and then extrapolates those judgments to the remaining document collection.” Maura R. Grossman and Gordon V. Cormack, “The Grossman-Cormack Glossary of Technology Assisted Review,” 7 Fed. Courts L. Rev. 1 (2013). If used correctly, TAR can result in drastic savings for clients in document-intensive cases and provide more accurate results. However, as the use of TAR expands, practitioners must be mindful of the relevant ethical pitfalls that may ensue.
Duty of Competence
Attorneys under ABA Model Rule 1.1 bear a duty to provide competent representation to their clients. Encompassed within this duty of competence is the duty to stay “abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology.” See ABA Model Rue 1.1, cmt. 8. Competence includes at least a basic understanding of TAR in order to evaluate the appropriate document review method for each case and implement a protocol for electronically stored information. More specifically, attorneys should be knowledgeable of the three general document-review methods: (1) linear review, i.e. reviewing each document one-by-one; (2) non-linear review, often involving concept searching, clustering, or near de-duplication; and (3) TAR, a variation on non-linear review. The specific facts and evidence of each case will dictate the appropriate method of review.
TAR is not a one-size-fits-all model. Electronic discovery vendors offer differing versions of TAR with their own respective software and coding methods. There are two general variations that attorneys should be aware of. The original version, which is traditionally referred to as predictive coding or TAR 1.0, is a method of document review in which attorneys analyze a subset of documents referred to as a “seed set” for relevance and then submit those decisions to a computer algorithm. The algorithm in turn: (1) examines the documents the attorneys have coded; (2) identifies similar documents within the collection; and (3) predicts whether the similar documents are relevant or non-relevant based on how the attorneys coded the seed set of documents.
The next generation of TAR, also known as continuous active learning or TAR 2.0, uses continuous active learning where no initial seed set is necessary. As the review team codes documents, the computer continuously takes and integrates the information and feeds the review team what it believes are the most responsive documents. When the review team gets to the point when the results are no longer relevant, the review has been completed. Attorneys must have at least a rudimentary understanding of the TAR methodology to competently advise clients and ultimately defend the use of TAR against any challenges raised by opposing counsel.
Communication with Clients Regarding TAR
A baseline level of competence regarding TAR goes hand in hand with an attorney’s obligation to discuss significant developments in the representation of clients. Model Rule 1.4 provides that a lawyer’s duty to communicate requires a lawyer to “reasonably consult with the client about the means by which the client’s objectives are to be accomplished” and to “explain a matter to the extent reasonably necessary to permit the client to make informed decisions.” In a document-intensive case, communications with a client should include a discussion of the advantages and disadvantages of using TAR. For example, advantages may include cost savings and efficiency that yields more accurate results whereas disadvantages might include controversies with opposing litigants that arise out of the use of TAR. Without at least a baseline understanding of TAR, lawyers will be unable to convey the risks and benefits of TAR to clients that would in turn allow clients to make informed decisions about the use of TAR.
Responsibilities Regarding Non-Lawyer Assistance
A corollary to the duty of competence in the TAR context is an attorney’s responsibilities regarding non-lawyer assistance. A lawyer may use non-lawyers outside the firm to assist the lawyer in rendering legal services to the client, including retention of an electronic-discovery vendor to assist with TAR. ABA Model Rule 5.3, cmt. 3. An attorney may ethically partner with an outside vendor but the onus is on the lawyer to ensure that the discovery vendor acts ethically. ABA Model Rule 5.3(b) (an attorney partnering with a vendor shall have “direct supervisory authority” over the “non-lawyer” and should ensure that the vendor’s conduct is “compatible with the professional obligations of the lawyer”). Ultimately, it is the attorney who bears the burden of not only ensuring that the electronic discovery vendor is competent but also understanding and supervising the vendor’s work in connection with the use of TAR.
TAR is an emerging field of the law that is developing far faster than jurisprudence can keep up. The implications of an attorney’s duties of confidentiality under Model Rule 1.6 is one area of TAR law that attorneys should keep abreast of. Attorneys and courts alike will be forced to address the issue of how much information a proponent of predictive coding should be required to provide before there is a risk of revealing client-confidential information. One context where this issue arises is whether a party is required to produce the “seed set” of documents used to train the TAR. Seed sets include irrelevant documents that could contain confidential and even privileged information that are not required to be turned over under discovery rules. Attorneys should also be mindful that privileged information might be inadvertently disclosed in connection with TAR. To protect against these concerns, attorneys should consider entering into protective orders, clawback agreements, and/or Rule 502(d) orders that mandate the return of privileged documents produced during discovery without parties fearing that the disclosure waives attorney-client or work-product privileges.
TAR can be used as a tool that can save clients both time and money. The recent revisions to the Federal Rules that emphasize proportionality as a factor that should be considered in limiting the scope of discovery leaves no doubt that TAR will eventually become commonplace in cases that involve large volumes of documents. Competent representation requires an attorney to have at least a baseline understanding of TAR. As the use of TAR in discovery becomes more prevalent, practitioners can count on courts holding both counsel and litigants to an ever-increasing ethical standard in connection with its use.