January 20, 2015
In our Mid-Year E-Discovery Update, we reported that 2014 was shaping up to be the "year of technology" in e-discovery. The remainder of the year more than lived up to those expectations.
Powerful new data analytics tools have become available for search and review, predictive coding pricing is becoming more accessible and its use appears to be gaining more traction, e-discovery technologies are becoming available through the cloud, and technologies that automate aspects of information governance are becoming increasingly available. New technologies are also creating new challenges, such as the increasing use of texting and other applications on smart phones and other mobile devices.
The important case law trends in 2014 have focused on (1) predictive coding and technology assisted review; (2) sanctions and the duties of counsel in connection with implementing legal holds; (3) text messaging and mobile devices; (4) proportionality; and (5) social media.
After a lengthy process that began in 2010, the Judicial Conference issued a final set of proposed amendments to the Federal Rules of Civil Procedure in September that, if approved by the Supreme Court, will become effective on December 1, 2015. The most significant proposed amendments deal with building in proportionality considerations to the scope of discovery, new requirements for responding to document requests, and uniform standards for imposing sanctions for preservation failures. Most significantly, the Judicial Conference substantially overhauled its proposed sanctions rule following the public comment period.
We also saw in late 2014 a greatly heightened interest emerge in information governance and the need to defensibly delete data for which there is no legal obligation or business need to retain. While information governance was primarily viewed in the past as a means of bringing high e-discovery costs under control, data breaches in 2014 that have resulted in highly publicized leaks of sensitive company emails and other confidential corporate information have added new urgency to the proper management and defensible deletion of data.
Looking forward to 2015 and beyond, we expect to see the following likely trends:
We invite you to review our discussion of these trends below, and to look out for our client updates, webinars and articles in the upcoming year.
In 2014, powerful new data analytics tools became available for search and review, predictive coding pricing became more accessible and its use appeared to be gaining more traction, e-discovery technologies became available through the cloud, and technologies that automate aspects of information governance began to emerge.
While we reported on these trends in our Mid-Year E-Discovery Update, we saw them continue during the remainder of 2014, along with an increasing recognition that quite often there is not a single "magic bullet" to solve the challenges of enormous document volumes, high review costs and often inflexible production deadlines.
Rather, various approaches can be taken to search and review, including a holistic approach that uses more than one methodology or technology at different stages of the process, or in combination with one another. The use of search terms has continued to be, by far, the predominant approach to search. But increasingly parties are combining the use of search terms with other technologies to improve the effectiveness and efficiency of their search and review processes.
Looking forward to 2015 and beyond, counsel who are literate in the use of the latest search and review technologies will realize substantial efficiencies and advantages in litigation for their clients. Additionally, in light of leaks of sensitive emails and other corporate information as a result of highly publicized hacks and thefts of terabytes of company data, we are likely to see a greater interest in information governance and the use of technological solutions–such as automated classification and disposal–to help ensure implementation of document retention and disposal policies.
A significant challenge, however, will be reconciling more complex approaches to search and review–often still using search terms as part of the process–with a judiciary and legal community that is very comfortable with the relatively straightforward, but often burdensome, methodology of running search terms through the document collection and having attorneys review every document that the search terms retrieve. Disputes related to the use of predictive coding, either alone or in combination with search terms, are an example of these growing pains.
In the past, the term "technology-assisted review" or "TAR" was often used as a synonym for predictive coding. While predictive coding has garnered much of the publicity about new search and review technologies, other technologies rightfully fall within the meaning of technology-assisted review. Indeed, in Chen-Oster v. Goldman, Sachs & Co., Magistrate Judge James Francis IV of the Southern District of New York even referred (correctly) to the use of search terms as a form of TAR. See Chen-Oster v. Goldman, Sachs & Co., 2014 WL 716521 (S.D.N.Y. Feb. 18, 2014), (holding that a party’s use of a technology-assisted review process does not mean that it must produce documents that the process identifies as potentially relevant without further human review).
One of the hallmarks of 2014 is that new visual analytics applications have become available that can be a powerful supplement to other search methodologies such as search terms and predictive coding. Such analytics applications organize documents based on their internal content and their similarity to other documents in the set. While concept clustering technologies have been available for a while, the new visual analytics applications provide a more powerful user interface that visually shows clusters of documents that users click through to more detailed levels, shows how the documents are related to one another, and shows their relationships with other clusters of documents. Once a relevant or important document is found, the tool can help find similar documents.
Visual analytics and other concept clustering applications can be used to find important documents early in a case and to supplement or confirm the effectiveness of other methodologies, such as search terms or predictive coding. See Gareth Evans and David Grant, Tools Let Attorneys Follow the Breadcrumbs, The National Law Journal (Sept. 1, 2014); see also David Grant, Seeing is Believing: Using Visual Analytics to Take Predictive Coding Out of the Black Box (FTI Technology White Paper).
Various analytics applications to improve the search and review process are not new, of course, but anecdotal evidence suggests they are being used more frequently than in the past. These include conceptual searching, using analytics to find documents similar to those already found, conceptual near duplicate detection, and clustering to group similar documents together into virtual folders displayed by topic. These tools can be used to find important documents more quickly and to code groups of documents more efficiently.
It has become popular of late to count the growing number of judicial decisions mentioning predictive coding and to interpret that as an indication of the increasing adoption of predictive coding as a search and review methodology. By that measure, 2014 was seemingly a banner year for predictive coding, with 17 cases mentioning predictive coding compared to 9 in 2013 and 6 in 2012).
But many of these cases, as in past years, merely involved references to the technology and not substantive discussions or approval of its use in the particular case. See, e.g., FDIC v. Bowden, No. 4:13-cv-245, 2014 WL 2548137 (S.D. Ga. Jun. 6, 2014) (ordering parties to "consider the use of predictive coding"); In re Domestic Drywall Antitrust Litig., 88 Fed. R. Serv. 3d 966 (E.D. Pa. 2014) (referring to the availability of predictive coding for searching ESI); Deutsche Bank Nat. Trust Co. v. Decision One Mortg. Co., LLC, No. 13 L 5823, 2014 WL 764707 (Ill. Cir. Ct. Jan. 28, 2014) ("If the parties agree that predictive coding would be appropriate in this case, they are encouraged to use that tool."); Aurora Coop. Elevator Co. v. Aventine Renewable Energy, No. 4:12-civ-230, slip op. at 1-2 (D. Neb. Mar. 10 2014) (ordering parties to "consult with a computer forensic expert to create search protocols, including predictive coding as needed, for a computerized review of the parties’ electronic records."); U.S. v. Exxonmobil Pipeline Co., No. 4:13-cv-00355, 2014 WL 2593781 (E.D. Ark. Jun. 10, 2014 (defendants suggested the use of predictive coding to ease the burden of production and to meet their production deadline, but the issue was not presented to the court for approval); Green v. Am. Modern Home Ins. Co., No. 1:14-cv-04074, 2014 WL 6668422 (W.D. Ark. Nov. 24, 2014) (entering parties’ stipulated ESI protocol allowing the parties with the option of using a "technology assisted review platform" in lieu of search terms to search documents); Good v. American Water Works, No. 2:14-cv-01374, 2014 WL 5486827 (S.D. W. Va. Oct. 29, 2014) (encouraging defendants to use technology-assisted review in combination with manual review to expedite privilege review).
Two cases reflected that plaintiffs’ counsel had used predictive coding in analyzing and reviewing documents they had received in document productions from defendants and third parties. In The New Mexico State Investment Council v. Bland, No. D-101-cv-2011-01434, 2014 WL 772860 (N.M. Dist. Feb. 12, 2014), the court in approving settlements in the litigation stated that "[i]n reviewing documents, [plaintiff’s counsel] implemented various advanced machine learning tools such as predictive coding, concept grouping, near-duplication detection and e-mail threading." Id. at *6. The court further stated that "[t]hese tools . . . enabled the reviewers on the document analysis teams to work more efficiently with the documents and identify potentially relevant information with greater accuracy than the standard linear review." Id. In approving a settlement and an award of attorney’s fees in Arnett v. Bank of America, No. 3:11-cv-1372, 2014 WL 4672458 (D. Or. Sept. 18, 2014), the court stated that plaintiff’s counsel reviewed the more than 1.1 million documents produced in the case using "search terms, predictive coding, and manual review methods." Id. at *9.
Similarly, three cases reflected that defendants had used predictive coding in reviewing documents for production in response to the initial round of document requests. In United States v. Univ. of Nebraska at Kearney, No. 4:11-cv-3209, 2014 WL 4215381 (D. Neb. Aug. 25, 2014), the defendant objected to subsequent document requests in light of its prior use of predictive coding. Id. at *3. Although the defendant argued for cost-shifting for the additional production, the court held that the additional document requests were overbroad and did not reach the cost-shifting issue. See id. at *5. In Smilovits v. First Solar, No. 2:12-cv-00555, slip op. at 1-2 (D. Ariz. Nov. 20, 2014), the court held that defendants’ use of predictive coding did not confine plaintiffs’ document discovery to the first round of requests and noted that defendants had not provided any information about the costs to "retrain" the predictive coding tool to deal with subsequent requests. And in In re Bridgepoint Educ., Inc. Sec. Litig., No. 12-cv-1737, 2014 WL 3867495 (S.D. Cal. Aug. 6, 2014), the court denied plaintiffs’ request to require the defendant to use predictive coding on custodians’ documents that it had previously searched using traditional search terms. Id. at *4.
In one case, the court ordered the defendant to use predictive coding to search more than 2 million documents after "little or no discovery was completed" before the discovery cutoff and the parties had ongoing disputes after "months of haggling" over search terms that yielded large numbers of documents for review. See Independent Living Center v. City of Los Angeles, No. 2:12-cv-00551, slip op at 1-2 (C.D. Cal. Jun. 26, 2014). The court also held that the defendant was not necessarily required to engage in a quality assurance process as part of the predictive coding protocol, but if the plaintiff insisted upon such a process, then plaintiff would be required to pay for 50% of its costs. Id. at 2-3.
A handful of cases addressed parties’ requests to use predictive coding over the objection of the opposing party. In Federal Housing Finance Agency v. JPMorgan Chase & Co., Inc., et al., No. 11-cv-6189, 2014 WL 584300, at *3 (SDNY Feb. 14, 2014), Judge Denise Cote indicated that she had approved the use of predictive coding over the objections of the plaintiff at an earlier hearing in the case. At that earlier hearing, Judge Cote is quoted in the transcript as having stated that "[i]t seems to me predictive coding should be given careful consideration in a case like this, and I am absolutely happy to endorse the use of predictive coding and to require that it be used as part of the discovery tools available to the parties." Id. (transcript of Jul. 24, 2012 hearing at 8).
In Progressive Casualty Ins. Co. v. Delaney, No. 2:11-cv-00678, 2014 WL 3563467, at *8-*11 (D. Nev. Jul. 18, 2014), Magistrate Judge Peggy Lean of the District of Nevada denied the plaintiff’s request to use predictive coding, primarily because the request was made extremely late in the discovery period and the plaintiff had previously agreed in the parties’ ESI protocol to use search terms and human review. Nevertheless, Judge Lean described in very positive terms the potential effectiveness of predictive coding and stated that she would not have hesitated to approve a predictive coding protocol had it been submitted earlier in the case. See id. at *8. The court also criticized the plaintiff’s plan to apply predictive coding to documents hitting the search terms and not to the entire document population. Id. at *10.
The Progressive decision is controversial, however, in that it also criticized plaintiff’s unwillingness to share with opposing counsel the irrelevant documents used to train the predictive coding tool. Judge Lean erroneously stated that, in the reported decisions that have approved predictive coding, "the courts have required the producing party to provide the requesting party with full disclosure about . . . the documents used to ‘train’ the computer." See id. at *10 (emphasis added) (citing Da Silva Moore and In re: Actos).
In both Da Silva Moore and Actos, which Judge Lean cited in Progressive (and in other cases allowing the opposing party to see irrelevant documents in the training sets), however, the parties seeking to use predictive coding had voluntarily stipulated to allow access to the irrelevant training documents. While in those matters the parties may not have been concerned about disclosing irrelevant documents to opposing counsel, that is often not the case–for example, in disputes among competitors, in disputes where the stakes are extraordinarily high, where there are concerns that counsel will use the knowledge gained from the irrelevant documents either in the present case or in the next case he or she brings (i.e., the "you can’t un-ring the bell" phenomenon), or where there are concerns that mistakes can be made in keeping the information confidential and the consequences of such disclosure are too great. Additionally, where document volumes are large and the prevalence of relevant documents is low, there can be several thousand irrelevant documents in the training and validation sets. In those circumstances, using predictive coding may be a non-starter if it means that the opposing party will get to see the irrelevant documents.
In contrast with Progressive, the court in Bridgestone Americas, Inc. v. IBM permitted the plaintiff to change its search and review methodology mid-stream to predictive coding. See Bridgestone Americas, Inc. v. IBM, No. 3:13-1196, 2014 WL 4923014, at *1 (M.D. Tenn. Jul. 22, 2014). The court also permitted plaintiff to undertake a hybrid approach, using predictive coding on documents initially identified through the use of search terms (nevertheless, resulting in over two million documents requiring review). The court expressly recognized that using predictive coding "is a judgment call" and that, in a case involving "millions of documents to be reviewed with costs likewise in the millions . . . . [t]here is no single, simple, correct solution possible under these circumstances." Id.
In Dynamo Holdings Ltd. Partnership v. Commissioner of Internal Revenue, 143 T.C. No. 9, 2014 WL 4636526 (Sept. 17, 2014), the tax court approved the petitioner’s use of predictive coding over the respondent’s objection that it is an "unproven technology." Addressing one of the issues that those considering using predictive coding frequently face–i.e., whether court approval is necessary–the court observed that the petitioner’s request to use predictive coding "is somewhat unusual." Id. at *3. The court stated that, "although it is a proper role of the court to supervise the discovery process and intervene when it is abused by the parties, the court is not normally in the business of dictating the process that they should use when responding to discovery." Id. "If our focus were on paper discovery," the court continued, "we would not (for example) be dictating to a party the manner in which it should review documents for responsiveness or privilege, such as whether that review should be done by a paralegal, a junior attorney, or a senior attorney." Id.
While stating that, if the respondent believes "the ultimate discovery response is incomplete," then it could file a motion to compel "at that time," the court nevertheless took up the issue of whether predictive coding would be allowed "because we have not previously addressed the issue of computer-assisted review tools[.]" Id. In response to the respondent’s assertion that predictive coding is an "unproven technology," the court stated that while predictive coding is a relatively new technique, "the understanding of e-discovery and electronic media has advanced significantly in the last few years, thus making predictive coding more acceptable in the technology industry than it may have previously been." Id. at *5. The court added that, "[i]n fact, we understand that the technology industry now considers predictive coding to be widely accepted for limiting e-discovery to relevant documents and effecting discovery of ESI without an undue burden." Id.
Whether these cases are the tip of a small or large iceberg–or something in between–of cases where predictive coding is being used is impossible to tell. By any measure, however, the number of reported decisions reflecting the use of predictive coding is very small compared to the overall number of cases being litigated. The increased number of decisions discussing predictive coding does appear to reflect an increase in awareness among the judiciary and litigants of predictive coding as an option for search and review. Some vendors of predictive coding technology claim that predictive coding is quietly gaining usage, and that it has particularly done so in 2014. Additionally, vendors’ pricing for predictive coding applications has generally come down quite significantly, making it more frequently a viable option from a cost perspective than in the past.
Regardless, it is apparent that predictive coding is still in its early stages and the existing jurisprudence has not finally resolved significant issues relating to its use. Those issues include (1) whether the use of predictive coding must be disclosed to the opposing party; (2) whether court approval is necessary; (3) whether predictive coding may be used in combination with keywords; (4) whether a party must disclose irrelevant documents in the seed, training and validation sets; (5) whether the review decisions on such documents are protected attorney work product; and (6) whether and under what circumstances a party may be compelled to use predictive coding when it has selected an alternative methodology.
Until there are more consistent and definitive rulings on these issues, or a general consensus emerges in the legal community, it is likely that the use of predictive coding will grow, but it will continue doing so at a slow pace.
Information governance–generally defined as an integrated approach to records management that involves legal, business operations and IT in managing, controlling and defensibly deleting data–has been a hot topic for the past three years. Information governance can help control the rising costs of e-discovery by helping to ensure that companies do not keep data that they have not legal obligation or business need to retain.
In 2014, data security became a much more prominent reason to implement an information governance program. Recent highly publicized leaks of sensitive company emails and other confidential corporate information in data breaches involving thefts of terabytes of company data have added urgency to the need to defensibly dispose of unneeded information.
While in the past many companies may have merely considered implementing information governance procedures or made only marginal changes in their existing practices, we expect that in 2015 there will be much more attention paid to actually undertaking information governance programs. Technologies such as automated document classification and disposal (think predictive coding for information governance) will also increasingly play a role in such programs. What remains to be seen is courts’ reaction to the very short retention periods that some companies are implementing.
As we reported in our Mid-Year Update, companies are also increasingly insourcing aspects of the e-discovery process that previously were handled by outside e-discovery vendors. One 2014 survey found that 90% of responding companies have internal teams, rather than an outside service provider, handle preservation and collection. See also Gareth Evans, Technology: Self-Collection is Not Always the Fox Guarding the Henhouse (Inside Counsel Dec. 27, 2013). Exceptions were noted for high-stakes matters where a third-party expert may need to testify on the defensibility of the collection process, matters where compliance with foreign data privacy regulations is an issue, and social media platforms and "bring your own device" environments involving more complex collection issues. Additionally, many companies are now handling data processing for e-discovery internally, which has been estimated to account for almost 20% of e-discovery costs. The availability of e-discovery software and storage through the cloud has also meant that more companies and law firms can handle internally some or all of the process that has traditionally been provided by outside e-discovery service providers.
While many companies are finding that preservation and collection may be well-suited to be managed in-house, they continue to outsource other aspects of the e-discovery process. The same survey, for example, found that 83% of respondents outsource review and production. Respondents cited staffing and internal budget limitations–as well as the expertise of outside providers–among the reasons for the ongoing need to outsource.
A series of opinions focusing on legal holds and the obligations of counsel in their implementation featured prominently in the e-discovery sanctions area in 2014.
In Brown v. Tellermate Holdings, Ltd., No. 2:11-cv-1122, 2014 WL 2987051 (S.D. Ohio July 1, 2014), the court imposed evidence preclusion and monetary sanctions for the defendant’s failure to timely preserve and produce relevant ESI in an age discrimination case. In doing so, the court focused on the duty of counsel to examine critically the client’s representations about the existence and availability of responsive documents.
The most significant failures the court found in Brown v. Tellermate related to a web-based application utilized by the defendant’s sales force, including the plaintiffs. See id. at *17-*20. When information from the application was requested, the defendant and its counsel asserted that the defendant was not in control of the data and could not produce it, which turned out to be untrue. Id. No steps had been taken to preserve the information based, at least in part, upon counsel’s "unfounded" and "mistaken" belief that a third-party service provider would preserve the data. See id. at *20. By the time counsel checked this assumption, the integrity of the data (which was ultimately produced) was in question because it had been subject to possible changes by the defendant’s sales force, which was still using the application. Id.
The court stated that "[l]ike any litigation counsel, Tellermate’s counsel had an obligation to do more than issue a general directive to their client to preserve documents which may be relevant to the case. Rather, counsel had an affirmative obligation to speak to the key players at Tellermate so that counsel and client together could identify, preserve, and search the sources of discoverable information." Id.
The court held that in this case counsel fell "far short" of their obligation to "examine critically" the information that Tellermate gave them about the existence and availability of documents the plaintiffs requested. See id. at *1. The court stated that counsel "have a duty (perhaps even a heightened duty) to cooperate in the discovery process; to be transparent about what information exists, how it is maintained, and whether and how it can be retrieved; and, above all, to exercise sufficient diligence (even when venturing into unfamiliar territory like ESI) to ensure that all representations made to opposing parties and to the Court are truthful and are based upon a reasonable investigation of the facts." Id. at *2.
Similarly, in Procaps S.A. v. Patheon Inc., No. 12-civ-24356, 2014 WL 800468 (S.D. Fla. Feb. 28, 2014), in an opinion invoking the Paul Newman film Cool Hand Luke ("what we’ve got here is a failure to communicate") and the band U2 ("but I still haven’t found what I’m looking for"), the court faulted plaintiff’s counsel in an antitrust suit for (i) failing to ensure that the plaintiff implemented a legal hold; (ii) failing to travel to Colombia (where the plaintiff is based) to meet with the plaintiff’s IT team or other executives to discuss how relevant ESI would be located; (iii) failing to retain an "ESI retrieval consultant" to help it implement a legal hold and to search for relevant ESI; and (iv) relying on the plaintiff’s own executives and employees to self-search for documents. See id. at *1-*2.
The court ordered the plaintiff to retain an e-discovery vendor to conduct an extensive analysis of its data sources, make forensic images of sources where potentially relevant ESI was located, identify custodians whose files must be searched, interview the custodians, and use agreed-upon search terms to search the documents. Id. at *3 –*5.
The court also imposed monetary sanctions, requiring that the plaintiff’s law firm be responsible for paying 50% of the sanctions. Furthermore, the court expressly urged–though it did not require–that the plaintiffs’ law firm identify "which attorneys caused, or helped cause, this discovery failure and to determine whether those attorneys (rather than the firm itself) should pay all or some" of the monetary sanctions imposed upon the law firm. Id. at *6.
In Knickerbocker v. Corinthian Colleges, 298 F.R.D. 670 (W.D. Wash. 2014), the court imposed monetary sanctions on the defendant and its counsel for a series of alleged failures, including (i) failure to issue a legal hold notice or to implement preservation procedures; (ii) reliance on employees to search for relevant documents and, compounding the problem, testimony from employees that they had not, in fact, been asked to search for relevant documents; (iii) allowing all of the plaintiff’s emails to be deleted from the defendant’s email server after commencement of the case pursuant to the company’s six-month retention period for emails; and (iv) counsel’s certification that all available sources had been searched despite the fact that the defendant had not searched its backup tapes. Id. at 678-81. Based on the court’s finding that both the defendant and its counsel’s conduct constituted bad faith, the court imposed monetary sanctions on both. Id. at 681-82.
There were other cases illustrating things that can go wrong in implementing legal holds. See, e.g., Vincente v. City of Prescott, No. CV-11-08204, 2014 WL 3894131 (D. Ariz. Aug. 8, 2014) (defendant issued legal hold notices but (i) did not coordinate with IT department, which failed to suspend autodelete for email; and (ii) relied upon individual custodians for self-collection); Osberg v. Foot Locker, No. 07–cv–1358, 2014 WL 3767033 (S.D.N.Y. July 25, 2014) (court held that the defendant failed to implement a legal hold until three years after suit was originally filed–although the suit was dismissed and later re-filed during that time–and that defendant conducted "spring cleanings" that destroyed potentially relevant documents).
With respect to the duties of counsel in handling e-discovery, the State Bar of California Standing Committee on Professional Responsibility and Conduct issued in 2014 its Formal Opinion Interim No. 11-0004 (ESI and Discovery Requests) for public comment. The interim opinion recognizes that while attorney competence related to litigation generally requires a basic understanding of and ability to handle issues relating to e-discovery, cases involving more complex e-discovery issues "may require a higher level of technical knowledge and ability[.]" See id.
The interim formal opinion states that "[a]ttorneys handling e-discovery should have the requisite level of familiarity and skill to, among other things, be able to perform (either by themselves or in association with competent co-counsel or expert consultants) the following: (1) initially assess e-discovery needs and issues, if any; (2) implement appropriate ESI preservation procedures, including the obligation to advise a client of the legal requirement to take actions to preserve evidence, like electronic information, potentially relevant to the issues raised in the litigation; (3) analyze and understand a client’s ESI systems and storage; (4) identify custodians of relevant ESI; (5) perform appropriate searches; (6) collect responsive ESI in a manner that preserves the integrity of that ESI; (7) advise the client as to available options for collection and preservation of ESI; (8) engage in competent and meaningful meet and confer with opposing counsel concerning an e-discovery plan; and (9) produce responsive ESI in a recognized and appropriate manner. Id.
According to the interim formal opinion, "[s]uch competency requirements may render an otherwise highly experienced attorney not competent to handle certain litigation matters involving ESI. An attorney lacking the required competence for the e-discovery issues in the case at issue has three options: (1) acquire sufficient learning and skill before performance is required; (2) associate with or consult technical consultants or competent counsel; or (3) decline the client representation. Lack of competence in e-discovery issues can also result, in certain circumstances, in ethical violations of an attorney’s duty of confidentiality, the duty of candor, and/or the ethical duty not to suppress evidence." Id.
Emails have traditionally been the primary source of relevant communications and documents. Increasingly, text messages and other data on mobile devices are becoming important and, not too surprisingly, the subject of sanctions decisions where a party has failed to preserve and collect relevant text messages.
The United States Supreme Court’s landmark decision in Riley v. California, 134 S. Ct. 2473 (2014), highlighted both the importance of mobile devices such as smart phones as a source of potentially relevant information and the privacy interests that can be involved in discovery of the information from such devices. In Riley, the Court unanimously held that the Fourth Amendment generally requires law enforcement to obtain a warrant before reviewing digital information that is stored on a smart phone seized incident to arrest. Id., 134 S. Ct. at 2485, 2493.
The Court observed that modern cellphones have the capacity to store "millions of pages of text, thousands of pictures or hundreds of videos" and thus "implicate privacy concerns far beyond those implicated by the search of a cigarette pack, a wallet, or a purse." See id. at 2478, 2488-89. The Court also stated that "it is no exaggeration to say that more than 90% of American adults who own a cellphone keep on their person a digital record of nearly every aspect of their lives." See id. at 2490.
The Court further stated that a cellphone’s immense storage capacity "has several interrelated consequences for privacy. First, a cellphone collects in one place many distinct types of information–an address, a note, a prescription, a bank statement, a video–that reveal much more in combination than any isolated record. Second, a cell phone’s capacity allows even just one type of information to convey far more than previously possible." Id. at 2478-79. The Court further recognized that "[a]lthough the data stored on a cell phone is distinguished from physical records by quantity alone, certain types of data are also qualitatively different." Id. at 2490. The Court referred to Internet search and browsing history that may reveal an individual’s private interests and concerns, such as "symptoms of disease, coupled with frequent visits to WebMD." Id.
While Riley dealt with the requirement of a warrant for law enforcement to search smart phones, the privacy interests that the Court recognized will likely lead to challenges in civil litigation to broad requests for information off of such devices. See, e.g., Bakhit v. Safety Marking, Inc., 2014 WL 2916490 at *2 (D.Conn. Jun. 26, 2014) (citing Riley in denying motion to inspect the data stored on individual defendants’ cell phones).
Now that mobile devices and the immense volumes of data they can store have become a feature of modern life, it is not surprising that data on such devices–in particular, text messages–have become the subject of sanctions decisions regarding alleged failures to preserve relevant information. In 2014, the Seventh Circuit upheld sanctions imposed in In Re Pradaxa (Dabigatran Etexilate) Prods. Liab. Litig., MDL No. 22385, 2013 WL 6486921 (S.D. Ill. Dec. 9, 2013), aff’d, 745 F.3d 216, 218 (7th Cir. 2014). In Pradaxa, the court held that the defendant had a duty to suspend an auto-delete function that operated on relevant text messages and imposed nearly $1 million in sanctions for having failed to do so on company-issued smart phones, among other things. The court found that the plaintiffs had expressly requested the text messages by including text messages in the boilerplate definition of "document," but the defendants failed to halt the auto-programmed delete function for text messages once a litigation hold was in place.
In Calderon v. Corporacion Puertorrique de Salud, 992 F. Supp. 2d 48, 52-53 (D.P.R. 2014), an employment discrimination case, the court held that an adverse inference instruction against the plaintiff was appropriate where the plaintiff had only selectively preserved relevant text messages between himself and a third-party. The court found that the plaintiff’s failure to preserve more than 38 text messages prejudiced the defendants by precluding a complete review of potentially relevant conversations and pictures sent via text messages. The court viewed the plaintiff’s actions as a "conscious abandonment of potentially useful evidence," indicating that "he believed those records would not help his side of the case." Id. at 52.
In Hosch v. BAE Systems Information Solutions, Inc., No. 1:13-cv-00825 (AJT/TCP), 2014 WL 1681694, at *2 (E.D. Va. Apr. 24, 2014), the district judge adopted the magistrate judge’s findings that the plaintiff had engaged in a series of intentional and bad faith discovery violations, including the permanent deletion of all text messages and voicemails, by wiping his iPhone just two days before turning it over to counsel. The court dismissed the plaintiff’s action with prejudice and awarded the defendant attorney’s fees and costs incurred in bringing motions to compel and a motion for sanctions.
In Ewald v. Royal Norwegian Embassy, No 11-cv-2116, 2014 WL 1309095 (D. Minn. Apr. 1, 2014), the court declined to impose spoliation sanctions against a defendant for failure to preserve the contents of a cell phone because the plaintiff failed to produce sufficient evidence of prejudice. Id. at *1. The plaintiff presented evidence relating to just one lost but potentially relevant text message, and the plaintiff had failed to pursue other avenues of discovery (e.g., deposition testimony) relating to that message or the existence of others. Id. at *2, *19.
We expect that decisions involving the discoverability of and duty to preserve text messages and other information on smart phones and other mobile devices will increase in the future. The defensibility of using "disappearing text" messaging apps that automatically delete text messages either immediately after they are read or within a period of time set by the sender also likely will be an issue that courts are called upon to address.
Limiting the scope of discovery based on proportionality interests has been a significant topic in e-discovery for several years, and 2014 was no exception. Several significant cases addressed proportionality in 2014.
In Lord Abbett Municipal Income Fund, Inc. v. Asami, No. C–12–03694, 2014 WL 5477639 (N.D. Cal. Oct. 29, 2014), the court invoked proportionality in granting the plaintiff’s motion that it not be required to continue to preserve 159 computers where there was no basis from which to reasonably conclude that the computers contained relevant evidence.
The court stated that Federal Rule of Civil Procedure 26(b)(2)(C)(iii) "sets forth a proportionality principle which requires courts to limit the frequency or extent of discovery where it determines that the ‘burden or expense of the proposed discovery outweighs its likely benefit, considering the needs of the case, the amount in controversy, the parties’ resources, the importance of the issues at stake in the action, and the importance of the discovery in resolving the issues.’" Id. at *3. Citing the Northern District of California’s Guidelines for the Discovery of Electronically Stored Information, the court further stated that "[t]his district recognizes that the proportionality principle applies to the duty to preserve potential sources of evidence." Id.
In United States v. Univ. of Nebraska at Kearney, No. 4:11-cv-3209, 2014 WL 4215381 (D. Neb. Aug. 25, 2014), the court denied a motion to compel production of documents in response to broad document requests. Even if it assumed that the requested documents might be relevant, the court stated that it must weigh the burden versus the benefit pursuant to Federal Rule of Civil Procedure 26(b)(2)(C)(iii). See id. at *5.
The court found that the additional costs of the proposed discovery would "far outweigh" what could be gained from it. Id. It rejected the government’s proposed broad search terms, such as "document* w/25 policy," as they would retrieve "thousands of documents that have no bearing on this case." Id. at *6. Additionally, the court rejected the government’s proposal that the defendant produce documents without review subject to a claw back agreement on the grounds that the review was necessary to protect the privacy interests of the students who were the subjects of the documents. Id.
In Kellogg Brown & Root Services, Inc. v. United States, 117 Fed. Cl. 1 (Jun. 26, 2014), an action against the federal government alleging breach of a contract to provide a broad range of support services for various Army operations in Iraq, the Court of Federal Claims granted a motion for protective order regarding a broad document request on proportionality grounds.
The document request sought "all documents related to" the government’s creation, receipt, circulation, exchange, response to, or use of an Army officer’s presentation on substandard wiring conditions (the presentation had already been produced). Id. at 6-7.
The court found that the request sought "potentially marginally relevant" information, yet would impose an enormous burden. Id. at 8. "This generic request–for ‘all documents’ having virtually any association or relation to the 2006 presentation–targets potentially massive quantities of material. The request fails to delineate any specific custodians by, inter alia, limiting the request to one or more individuals, or even specific government entities. Rather, the request encompasses all government entities, and all of their current and former employees who served at all levels of government (ministerial to senior executive). Moreover, the request lacks any temporal limitation and, thus, requires the government to search, and potentially produce, at least eight to ten years of files in order to capture materials related to the presentation’s initial creation in 2006 through to its present use." Id.
Additionally, the court stated that the request "would require the government to invade an overwhelming number of computers, email accounts, and paper files from both past and present government employees and would require the government to devote substantial time and resources to the review of the material for relevancy, privilege, and security." Id. Citing the proportionality principle of Rule of Civil Procedure 26(b)(2)(C)(iii), the court granted the motion for a protective order with respect to the document request. Id.
With the proposed amendments to the Federal Rules of Civil procedure moving proportionality from its current place as a potential limit on discovery in Rule 26(b)(2)(C)(iii) to the definition of the scope of discovery itself in Rule 26(b)(1), we can expect frequent litigation of proportionality issues in the future.
The number of cases focusing on the discovery of social media continued to skyrocket in the second half of 2014. Reflecting that the use of social media continues to proliferate in business and social contexts, courts and commentators alike have noted that discovery of social media is now routine.
As reflected by the volume of reported decisions relating to social media, courts are still struggling to develop rules and protocols applicable to social media evidence, including whether special authentication rules should govern social media evidence, what threshold showing of relevance must be made before discovery of personal social media data should be allowed, and when the duty to preserve social media evidence arises. Bar associations have jumped into the fray to offer guidance to attorneys regarding the ethical duties implicated by preservation of clients’ social media evidence.
As explained in our Mid-Year Update, courts have taken two basic approaches to the authentication of social media evidence, which the Delaware Supreme Court coined "the Maryland approach" and "the Texas approach." Parker v. State, 85 A.3d 682, 684 (Del. 2014).
Under the Maryland approach, there are three permissible methodologies for authenticating social media evidence: "the testimony of the creator, documentation of the internet history or hard drive of the purported creator’s computer, or information obtained directly from the social networking site." Id. at 683 (citing Griffin v. State, 19 A.3d 415 (Md. 2011)). Unless the proponent can "convince the trial judge that the social media post was not falsified or created by another user" via one of these methods, the evidence "will not be admitted and the jury cannot use it in their factual determination." Id.
Conversely, under the Texas approach, "a proponent can authenticate social media evidence using any type of evidence so long as he or she can demonstrate to the trial judge that a jury could reasonably find that the proffered evidence is authentic." Id. (citing Tienda v. State, 358 S.W.3d 633 (Tex. Crim. App. 2012)).
The Second Circuit recently considered the issue, noting that "[s]ome courts have suggested applying ‘greater scrutiny’ or particularized methods for the authentication of evidence derived from the Internet due to a ‘heightened possibility for manipulation.’" U.S. v. Vayner, 769 F.3d 125, 131 n.5 (2d Cir. 2014) (citing Griffin). The court went on to hold that, "[a]lthough we are skeptical that such scrutiny is required, we need not address the issue as the government’s proffered authentication in this case falls under Rule 901’s general authentication requirement." Id.
In Vayner, the defendant was accused of transferring a false identification document. Id. at 127. The primary evidence submitted by the state was testimony from a witness who said he had received the document from defendant via a particular email address, and a profile page from a Russian social networking site (VK.com), which included a variation of defendant’s name, a photo of defendant, two places of employment where defendant had allegedly worked in the past, and a skype moniker that matched the moniker contained in the email address alleged to have been used to transfer the false document. Id. at 127-28.
The district court found the document to be authentic, noting "The information on there, I think it’s fair to assume, is information which was provided by [the defendant]," and "There’s no question about the authenticity of th[e] document so far as it’s coming off the Internet now." Id. at 128. A Special Agent with the State Department’s Diplomatic Security Service then testified regarding the content of the profile page, admitting on cross-examination he only had "cursory familiarity" with VK.com and did not know whether any identity verification was required in order for a user to create an account. Id. at 128-29.
The Second Circuit reversed, finding the district court abused its discretion in admitting the social networking profile page "because the government presented insufficient evidence that the page was what the government claimed it to be–that is, [defendant’s] profile page, as opposed to a profile page on the Internet that [defendant] did not create or control." Id. at 127. The court compared the profile page to "a flyer found on the street that contained [defendant’s] Skype address and was purportedly written or authorized by him," and reasoned "the district court surely would have required some evidence that the flyer did, in fact, emanate from [defendant]." Id. at 132. The court vacated the conviction and remanded the case for a new trial, finding the error not to be harmless. Id. at 134.
Because the court "express[ed] no view on what kind of evidence would have been sufficient to authenticate" the profile page, it is unclear whether in application the court would require the high bar of the Maryland approach, the lower bar of the Texas approach, or something in between. Id. at 133.
The Vayner decision follows a general trend wherein courts have found the testimony of the individual who printed the webpage in question to be insufficient to authenticate social media evidence. See, e.g., Moroccanoil, Inc. v. Marc Anthony Cosmetics, Inc., — F.Supp.3d —-, No. CV 13–2747 DMG (AGRx), 2014 WL 5786253, at *7 n.5 (C.D. Cal. Sept. 16, 2014) ("Defendant’s argument, that [Facebook screenshots] could be ‘authenticated’ by the person who went to the website and printed out the home page, is unavailing. It is now well recognized that ‘Anyone can put anything on the internet.’ [citations omitted] No website is monitored for accuracy.")
Courts continue to hold that "the fact that the information [sought] is in an electronic file as opposed to a file cabinet does not give [the party seeking discovery] the right to rummage through the entire file." Del Gallo v. City of New York, 43 Misc.3d 1235(A), at *6 (N.Y. Sup. Ct. 2014) (internal citations and quotations omitted). As with more traditional forms of evidence, the party seeking discovery "must establish a factual predicate for their request by identifying relevant information in [the social media] account — that is, information that contradicts or conflicts with plaintiff’s alleged restrictions, disabilities, and losses, and other claims." Id. at *3. (internal citations and quotations omitted). Absent such a showing, "granting carte blanche discovery of every litigant’s social media records is tantamount to a costly, time consuming fishing expedition." Id. at *5. (internal citations and quotations omitted)
In DelGallo, defendants argued that because plaintiff claimed to be "totally disabled" as a result of the accident in question, they were entitled to access all of plaintiff’s social media sites to learn about plaintiff’s post-accident condition and her self-assessment of the extent of her injuries. Id. at *7-8. The court ordered plaintiff to produce information from her LinkedIn account related to job offers, inquiries, and searches because such information was relevant to damages, but denied defendant’s request to access any other social media information based on "the mere hope of finding relevant evidence." Id.
Other courts have similarly denied discovery of the entirety of social media accounts because of an insufficient threshold showing of relevancy to the claims at issue. See, e.g., Finkle v. Howard County, Md., 2014 WL 6835628, at *1-2 (D. Md. Dec. 02, 2014) (Gallagher, Mag. J.) (denying as "overly broad" and "not reasonably calculated to lead to the discovery of admissible evidence" plaintiff’s requests for all personal email and social networking account information for nonparty employees involved in alleged discrimination); Doe v. Rutherford County, Tenn., Bd. of Educ., No. 3:13–0328, 2014 WL , at *1-3 (M.D. Tenn. August 18, 2014) (Bryant, Mag. J.) (granting defendant social media discovery from certain plaintiffs whose public social networking profiles included relevant information, but denying social media discovery from other plaintiffs for which there was no predicate showing); Stonebarger v. Union Pac. Corp., No. 13–cv–2137–JAR–TJJ., 2014 WL 2986892, at *2-5 (D. Kan. July 2, 2014) (James, Mag. J.) (denying defendant’s request for blanket access to plaintiff’s social networking profiles, but "allowing defendant to discover information relevant to plaintiff’s emotional state (which plaintiff put at issue)," which protected "plaintiff from a fishing expedition"); Smith v. Hillshire Brands, No. 13–2605–CM, 2014 WL 2804188, at *3-6 (D. Kan. June 20, 2014) (O’Hara, Mag. J.) ("allow[ing] defendant to discover not the contents of plaintiff’s entire social networking activity, but any content that reveals plaintiff’s emotions or mental state, or content that refers to events that could reasonably be expected to produce in plaintiff a significant emotion or mental state").
A continuing theme in 2014 was the extent to which parties have an obligation to preserve social media during litigation, and whether the modification of social media constitutes sanctionable spoliation. Because social media is dynamic, account holders may delete information from their page or cancel their account altogether, without realizing that the information could be relevant to an anticipated or pending matter.
In determining whether to award sanctions for spoliation of social media, courts have focused on whether it was reasonably foreseeable that the information would be sought in discovery, and whether the users had a duty to preserve their account at the time the evidence was deleted.
In Keller v. Keller, the court found that because its order to preserve all electronically stored information (ESI) did not specifically include social media information, the plaintiff’s changes to her Facebook page were not a willful violation of the order, reasoning, "Although the defendant’s motion included social media specifically, the court’s order did not. It is possible to successfully argue that a social media account such as Facebook is a form of ESI, but it would not rise to the unambiguous level of understanding necessary to make a finding of contempt for a violation of such an order." No. MMXFA114014330S, 2014 WL 4056926, at *8-9 (Conn. Super. Ct. July 9, 2014).
The court found that changes plaintiff made to her Facebook account after the discovery special master entered an order to preserve ESI and specially extended it to include social media sites, however, were "willful contempt of a clear and unambiguous order." Id. at *8. See also Painter v. Atwood, No. 2:12–CV–1215 JCM (NJK), 2014 WL 3611636, at *2 (D. Nev. July 21, 2014) (upholding magistrate judge’s finding of spoliation sanctions against plaintiff who deleted social media posts and text messages that defendant argued contradicted her sexual harassment claims).
Additionally, the ethics committees of various bar organizations have begun to weigh in on the duty of attorneys to advise their clients regarding the preservation of social media. In July, the Philadelphia Bar Association issued an opinion that indicated an attorney may instruct a client to delete damaging information from a social media account, but must also "take appropriate action to preserve the information in the event it should prove to be relevant and discoverable." Philadelphia Bar Ass’n Prof’l Guidance Comm., Op. 2014-5 (July 14, 2014). Also in July, the North Carolina State Bar issued Formal Ethics Opinion 5, which found a lawyer "may advise a client to remove information on social media if not spoliation or otherwise illegal." See North Carolina State Bar 2014 Formal Ethics Op. 5, July 25, 2014.
In September, the Pennsylvania Bar Association followed suit, issuing Formal Opinion 2014-13, which allows an attorney to advise a client to take down damaging material from social media, subject to spoliation concerns. Pennsylvania Bar Ass’n Formal Op. 2014-13. See also New York County Lawyers Ass’n Ethics Op. 745 (2013) ("Provided that such removal does not violate the substantive law regarding destruction or spoliation of evidence, there is no ethical bar to ‘taking down’ such material from social media publications, or prohibiting a client’s attorney from advising the client to do so, particularly inasmuch as the substance of the posting is generally preserved in cyberspace or on the user’s computer.").
After a drawn-out process that began in 2010, the proposed e-discovery amendments to the Federal Rules of Civil Procedure took a major step in 2014 towards actual enactment. Following the closure of a several-month public comment period, the Judicial Conference of the United States, the federal judiciary’s policymaking body, approved a final set of proposed rule amendments on September 16, 2014.
Nevertheless, we still won’t see any of the proposed rule amendments taking effect until nearly the end of 2015. The United States Supreme Court must approve the proposed amendments and then, barring Congressional intervention, they will take effect on December 1, 2015.
The affected rules include Rules 1, 4, 16, 26, 30, 31, 33, 34 and 37(e). Of particular note, the Judicial Conference in its last iteration of the proposed amendments dropped proposed limits on the number of document requests, interrogatories and requests for admission. It also substantially changed the proposed sanctions rule for failures to preserve relevant information. The following is a summary of the key proposed amendments affecting e-discovery.
The proposed amendment to Rule 1 was originally crafted to require cooperation among the parties, but that language was dropped relatively early on out of concerns that it would only spawn tangential disputes regarding whether parties were being sufficiently cooperative.
In its final proposed form, the language of Rule 1 would be altered to provide that the rules of civil procedure would not only be "construed and administered" (the current language) but also "employed by the court and the parties" to secure the just, speedy, and inexpensive determination of every action and proceeding.
The concept of cooperation still made it into the Committee Note that accompanies the proposed amendment, which states that "effective advocacy is consistent with – and indeed depends upon – cooperative and proportional use of procedure."
The proportionality factors currently listed in Rule 26(b)(2)(C)(iii) would be moved, with certain modifications, to the scope of discovery provision of Rule 26(b)(1).
As amended, Rule 26(b)(1) would permit a party to obtain discovery regarding any non-privileged matter that is relevant to any party’s claim or defense "and proportional to the needs of the case, considering the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit." Moving the language from a provision providing a limitation on the scope of discovery to the definition of the scope of discovery itself is viewed as a means of strengthening proportionality in discovery.
The amendments also would remove from current Rule 26(b)(1) the language that relevant information need not be admissible at trial if it is reasonably calculated to lead to the discovery of admissible evidence. According to the Committee Note, that language "has been used by some, incorrectly, to define the scope of discovery."
Previously proposed presumptive limits on the number of depositions (from the current 10 to 5), interrogatories (from the current 25 to 15), and requests for admission (from no current limit to 25 (except for authentication of documents)) have been withdrawn by the Judicial Conference and are no longer part of the rules amendment package.
The proposed amendment to Rule 34 requires that objections to document requests must be "state[d] with specificity" and also requires that the responding party state in its written responses "whether any responsive materials are being withheld on the basis of [an] objection." According to the Committee Note, this requirement "should end the confusion that frequently arises when a producing party states several objections and still produces information, leaving the requesting party uncertain whether any relevant and responsive information has been withheld on the basis of the objections."
The Committee Note clarifies that the responding party is not required to "provide a detailed description or log of all documents withheld." Rather, it states that the responding party must "alert other parties to the fact that documents have been withheld and thereby facilitate an informed discussion." It further provides that an objection "that states the limits that have controlled the search for responsive and relevant materials qualifies as a statement that the materials have been ‘withheld.’"
The Judicial Conference substantially revised its proposed amendment to Rule 37(e) following widespread criticism of aspects of the prior proposed amendment.
The proposed amendment in its latest iteration would replace the current language of Rule 37(e) with a rule would permit a court to impose curative measures or sanctions if ESI that should have been preserved "is lost because a party failed to take reasonable steps to preserve it, and it cannot be restored or replaced through additional discovery[.]"
The proposed sanctions rule consists of several important elements: (1) For curative measures or sanctions to be imposed, a party must have failed to have taken reasonable steps to preserve the lost information. (2) Curative measures or sanctions can only be imposed if the lost information cannot be restored or replaced through additional discovery. (3) A court may order curative measures only "upon finding prejudice to another party from the loss of the information." (4) The proposed rule contains an explicit proportionality requirement–i.e., the measures ordered must be "no greater than necessary to cure the prejudice." (5) the most severe sanctions–evidentiary estoppel, an adverse inference instruction, or case terminating sanctions–may be ordered "only upon a finding that the party acted with intent to deprive another party of the information’s use in the litigation."
The Civil Rules Advisory Committee of the Judicial Conference intended the proposed amendment to Rule 37(e) to resolve a circuit split regarding when these most severe of sanctions are appropriate, as it allows imposition of the most severe sanctions only when a party acts intentionally. See Report of the Advisory Committee on Civil Rules, at 50. C.f. Residential Funding Corp. v. DeGeorge Financial Corp., 306 F.3d 99 (2d Cir. 2002) (permitting severe sanctions for negligent conduct in the Second Circuit). The Civil Rules Advisory Committee endorsed the higher standard for sanctions because it believed that the negligence standard often goes beyond restoring the evidentiary balance in a case, it can be too harsh considering the ease with which ESI is lost, and can lead to costly over-preservation.
Given the December 1, 2015 effective date for the proposed rule amendments, we likely will not know until at least mid-2016 (or later) whether the amendment to Rule 37(e) appears to be fulfilling these goals in practice and can bring more fairness to the e-discovery sanctions area. In other words, don’t expect amended Rule 37(e) to have a significant impact any time soon.
The developments discussed above demonstrate, in our view, that 2014 was indeed "the year of technology" in e-discovery. While new technologies help create solutions to long-standing challenges, they also create new complexities–such as the proliferation of data sources, the ubiquitous use of mobile devices and applications that run on them (in particular, text messaging), and determining what technologies to use and how to use them in search and review. Obtaining advice from counsel and technical personnel who are well versed in e-discovery technologies and the legal and practical issues associated with them is more important than ever.
Gibson Dunn will continue to track the latest developments and trends. Please look for our updates and our attorneys’ articles, and for our 2015 Mid-Year E-Discovery Update, which will be published early in the third quarter of 2015.
Gibson Dunn & Crutcher’s lawyers are available to assist in addressing any questions you may have regarding the issues discussed in this update. The Electronic Discovery and Information Law Practice Group brings together lawyers with extensive knowledge of electronic discovery and information law. The group is comprised of seasoned litigators with a breadth of experience who have assisted clients in various industries and in jurisdictions around the world. The group’s lawyers work closely with the firm’s technical specialists to provide cutting-edge legal advice and guidance in this complex and evolving area of law. For further information, please contact the Gibson Dunn lawyer with whom you usually work or the following Co-Chairs of the Electronic Discovery and Information Law Practice Group:
© 2015 Gibson, Dunn & Crutcher LLP