THE AMERICAN LAW INSTITUTE Continuing Legal Education. Advanced Employment Law and Litigation February 27 March 1, 2014 Washington, D.C.

Transcription

1 1177 THE AMERICAN LAW INSTITUTE Continuing Legal Education Advanced Employment Law and Litigation 2014 February 27 March 1, 2014 Washington, D.C. Electronic Discovery Problems in Employment Litigation By Lynne Bernabei Alan R. Kabat Bernabei & Wachtel, PLLC Washington, D.C.

2 1179 ELECTRONIC DISCOVERY PROBLEMS IN EMPLOYMENT LITIGATION This article discusses several cutting-edge issues in electronic discovery in employment litigation: (1) the use of keyword searching versus predictive coding for reviewing and producing large sets of documents; (2) privilege and waiver issues arising from electronic discovery, including the impact of Rule 502, Fed. R. Evid.; and (3) ethical issues arising from accessing another party s s or social media sites. I. Predictive Coding and Keyword Searches for Voluminous Documents. In the dark ages only a decade or so ago the customary practice in reviewing documents in order to identify the relevant, responsive, and non-privileged documents for production was to have an attorney or paralegal review each page of each document to determine whether (1) it should be produced, or (2) withheld in part or entirely on the grounds of privilege, or (3) not produced as not responsive. As the volume of documents grew exponentially, thanks to the proliferation of , attorneys and their clients realized that this manual, page-by-page search, could be expensive and time-consuming, particularly in class actions that might involve hundreds or thousands of employees working in multiple offices. Hence, attorneys started doing keyword searches and searches limited to certain record custodians. The parties would agree to limit the document searches to those in the electronic files and s of a defined subset of employees ( custodians ), and that those documents would be electronically searched by keywords. Thus, only the documents that came up through a keyword search would have to be reviewed for privilege, with the presumption being that any document with one or more keywords would be relevant and responsive, while documents without any keywords would not be relevant or responsive and need not be considered any further. Keyword searching, of course, assumes that the parties can reach agreement on the keywords and the record custodians. Keyword searching can also be combined with Boolean searching, which is similar to searches done in Westlaw or Lexis, where one uses logical connections (e.g., and or w/10 ) to search for court decisions containing certain words or phrases, including those are close to each other, as opposed to court decisions in which those words are widely separated. For example, in employment litigation, the records custodians could be defined to include the following: (1) plaintiff(s); (2) the supervisors and other employees involved in the alleged adverse employment actions; (3) comparators; and (4) co-workers who are known or believed to have witnessed the employment actions. Similarly, keywords in employment litigation would include not only the names of the plaintiff(s) and any individual defendant(s), in order to capture s in which other people 1

3 1180 discussed the parties, but also words relating to important terms in the complaint and the answer. For example, in a case involving discrimination relating to bonus payments, keywords would include the name of the company s bonus or incentive program, as well as the terms used for the targets in that program. In a case involving discrimination based on a failure to provide comparable training opportunities to all employees, keywords would include words and phrases relating to the company s training programs and the selection process for those programs. The parties can also conduct staged searches, starting with a subset of keywords, and using the results to identify further keywords to be used, as a way of more precisely focusing the search. However, there has been an increasing recognition that keyword searching is not a panacea. It is both overinclusive and underinclusive. For example, searching for s with an employee s name may yield s about other employees with the same first and/or last name overinclusive. Conversely, searching for s in which stereotyped remarks are made will not yield s that use phrases that the attorneys did not think of including in the keywords underinclusive. These problems led U.S. Magistrate Judge Andrew Peck to conclude that keyword searching is not the last word in electronic discovery: In too many cases, however, the way lawyers choose keywords is the equivalent of the child s game of Go Fish. The requesting party guesses which keywords might produce evidence to support its case without having much, if any, knowledge of the responding party s cards (i.e., the terminology used by the responding party s custodians). Indeed, the responding party s counsel often does not know what is in its own client s cards. Da Silva Moore v. Publicis Groupe S.A., 287 F.R.D. 182, 191 (S.D.N.Y. 2012), aff d, 2012 WL (S.D.N.Y. Apr. 26, 2012). 1 Similarly, U.S. District Judge Shira Scheindlin recently concluded that: There is increasingly strong evidence that keyword searching is not nearly as effective at identifying relevant information as many lawyers would like to believe. National Day Laborer Organizing Network v. U.S. Immigration & Customs Enforcement Agency, 877 F. Supp. 2d 87, 109 (S.D.N.Y. 2012) (quoting M.R. Grossman & T. Sweeney, What Lawyers Need to Know About Search Tools: The Alternatives to Keyword Searching Include Linguistic and Mathematical Models for Concept Searching, National Law Journal (Aug. 23, 2010)). Thus, the latest wave in e-discovery is predictive coding, also known as computerassisted coding. Magistrate Judge Peck, who authored the first reported decision on predictive 1 The plaintiffs in Da Silva Moore, dissatisfied with Magistrate Judge Peck s rulings, sought to disqualify him based on the fact that he wrote an article about predictive coding, so that he was allegedly biased in favor of predictive coding. The court denied their motion, see 868 F. Supp. 2d 137 (S.D.N.Y. 2012), and the Second Circuit denied their mandamus petition. In re Da Silva Moore, No (2d Cir. Apr. 10, 2013) (per curiam). 2

4 1181 coding, noted that most lawyers already use predictive coding without being aware of it the spam filters in programs use predictive coding to segregate s that are presumed to be spam, i.e., pornographic s, advertisements and solicitations from sources known or likely to be scams, and other s of questionable veracity. Da Silva Moore, 287 F.R.D. at 184 n.3 (spam filters use predictive coding). Unfortunately, attorneys are well aware that spam filters are both overinclusive and underinclusive legitimate client sometimes gets trapped in the spam filters, while true spam shows up in inboxes. Nonetheless, predictive coding is here and employment lawyers need to prepare for it. Magistrate Judge Peck provided a general definition of this process: By computer-assisted coding, I mean tools (different vendors use different names) that use sophisticated algorithms to enable the computer to determine relevance, based on interaction with (i.e., training by) a human reviewer. Unlike manual review, where the review is done by the most junior staff, computerassisted coding involves a senior partner (or [small] team) who review and code a seed set of documents. The computer identifies properties of those documents that it uses to code other documents. As the senior reviewer continues to code more sample documents, the computer predicts the reviewer s coding. (Or, the computer codes some documents and asks the senior reviewer for feedback.) When the system s predictions and the reviewer s coding sufficiently coincide, the system has learned enough to make confident predictions for the remaining documents. Typically, the senior lawyer (or team) needs to review only a few thousand documents to train the computer. Some systems produce a simple yes/no as to relevance, while others give a relevance score (say, on a 0 to 100 basis) that counsel can use to prioritize review. For example, a score above 50 may produce 97% of the relevant documents, but constitutes only 20% of the entire document set. Counsel may decide, after sampling and quality control tests, that documents with a score of below 15 are so highly likely to be irrelevant that no further human review is necessary. Counsel can also decide the cost-benefit of manual review of the documents with scores of Da Silva Moore, 298 F.R.D. at (quoting A. Peck, Search, Forward, L. Tech. News, Oct. 2011, at 25, 29); see also M. R. Grossman & G. V. Cormack, Glossary of Technology-Assisted Review, 7 Federal Courts L. Rev. 1 (2013) (technical definitions of the terms used in this field) 3

5 1182 (online at: How exactly does predictive work, putting aside proprietary software methods? The following description is taken from Da Silva Moore, an employment case brought against a large advertising agency, in which the named plaintiffs sought to bring both a class action and a collective action under federal and New York anti-discrimination statutes. The employer had some three million electronic documents that had to be reviewed for responsiveness, relevancy, and privilege. Defendant, over plaintiffs objection, insisted on pursuing predictive coding, thus resulting in the first reported opinion in this field. First, the parties must agree upon record custodians, temporal scope of discovery, and upon the broad categories, keywords, or issues to be used to identify relevant documents. Then, the producing party selects a random sample (e.g., several thousand documents) of the entire and other electronic files associated with those record custodians. An experienced attorney for the producing party, familiar with the claims and defenses in the lawsuit, then reviews that subset, and codes them as to whether they are relevant, and as to which of the issue(s) are associated with each document. The result is that each document in the subset is scored with a code (on a scale) as to relevancy. The attorney also sets aside documents that are privileged. The software program keeps track of how each document is coded by the attorney. Based upon that data, the program uses proprietary technology to develop algorithms for analyzing the pool of documents to identify and code the most relevant documents. The producing party then turns over the non-privileged subset of the reviewed documents, together with the code for each such document, and allows the receiving party to do its own predictive coding analysis of that subset of documents. If the parties reach significantly different results as to what is relevant, they can discuss the coding process, and do successive iterations, until both sides are satisfied with the coding of each document in that subset. The parties also have to agree on the cut-off level for the codes, i.e., the number on the scale from 0 to 100 for which documents below that cut-off will not be produced. Then, the software program takes over based on its collective analysis of the attorneys coding of the subset, the software then runs through the remaining thousands or millions of documents to code them. The producing party then only has to do a privilege review of the documents above the cut-off level, in order to redact or withhold entirely any privileged documents. Critically, documents that are below the cut-off level will not be reviewed by anyone, since they are presumptively not relevant or responsive. 4

6 1183 It is obvious that there can be disputes at each stage of this process the identification of the record custodians; the temporal scope of discovery; the identification of issues and keywords to be used in coding documents for relevancy; the size of the random sample to be analyzed by the attorney; the numerical codes to be assigned to any given document; the number of iterative reviews, if any; and the cut-off score to be used. Indeed, in Da Silva Moore, the parties disputed each of these issues. Da Silva Moore, 287 F.R.D. at Commentators noted that in some cases, the inability of parties to agree on predictive coding has brought discovery to a standstill and forced courts to weed through infinite discovery motions. See P. Nair & L.N. Tookey, Five Cases Made for a Momentous Year in Predictive Coding, National Law Journal, Jan. 28, 2013, at 10, 11. Magistrate Judge Peck noted that his decision does not mean [that] computer-assisted review must be used in all cases, or that the exact ESI protocol approved here will be appropriate in all future cases that use computer-assisted review. 287 F.R.D. at 193. However, the takeaway lesson is that computer-assisted review is an available tool and 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 for keywords or any other technological solution to ediscovery, counsel must design an appropriate process, including use of available technology, with appropriate quality control, to review and produce relevant ESI while adhering to [Fed. R. Civ. P.] Rule 1 and Rule 26(b)(2)(C) proportionality. Id. Judge Scheindlin similarly endorsed the potential use of predictive coding in addition to (or instead of) keyword searching: And beyond the use of keyword search, parties can (and frequently should) rely on latent semantic indexing, statistical probability models, and machine learning tools to find responsive documents. Through iterative learning, these methods (known as computer-assisted or predictive coding) allow humans to teach computers what documents are and are not responsive to a particular FOIA or discovery request and they can significantly increase the effectiveness and efficiency of searches. National Day Laborer, 877 F. Supp. 2d at 109. However, not all judges have treated predictive coding as worthwhile or necessary. In a medical products liability case, U.S. District Judge Robert Miller of the Northern District of Indiana rejected the plaintiffs attempt to require the defendant to undertake predictive coding searches of its electronic documents. In re Biomet M2A Magnum Hip Implant Products Liability Litig. (MDL 2391), No. 3:12-MD-2391, 2013 WL (N.D. Ind. Apr. 18, 2013). Biomet 5

7 1184 had already produced some 2.5 million documents, but the plaintiffs believed that the production should run something closer to 10 million documents, id. at *1, so plaintiffs demanded that Biomet go back to its 19.5 million documents and employ predictive coding. Id. at *2. Here, Biomet s e-discovery costs [already] are about $1.07 million and will total between $2 million and $3.25 million. Id. at *1. Judge Miller found it unnecessary to address plaintiffs arguments that predictive coding is a better way of doing things than keyword searching (although the statistics as to predictive coding were not as well documented as plaintiffs asserted), id. at *2, since his role was instead to decide whether Biomet s procedure satisfies its discovery obligations, and he found that what Biomet has done complies fully with the requirements of Federal Rules of Civil Procedure 26(b) and 34(b)(2). Id. Requiring Biomet to go back to the original pool of 19.5 million documents, and do a new review was unjustified, particularly since it would cost Biomet a million, or millions, of dollars to test the [plaintiffs ] theory that predictive coding would produce a significantly greater number of relevant documents. Id. at *3. Thus, the benefit of this discovery did not equals or outweighs its additional burden on, and additional burden on, and additional expense to, Biomet under Rule 26(b)(2)(C), Fed. R. Civ. P. Id. Only if plaintiffs would agree to bear the expense for production of documents that can be identified only through re-commenced processing, predictive coding, review, and production, would Biomet have to use predictive coding. Id. In a subsequent decision, the plaintiffs claimed that they needed to know exactly which documents Biomet used in the training of the predictive coding algorithm, since without knowing the training documents, the [plaintiffs ] Steering Committee says, it can t intelligently propose more search terms, since it doesn t know what already has been included in the search. In re Biomet M2A Magnum Hip Implant Products Liability Litig. (MDL 2391), No. 3:12-MD- 2391, 2013 WL , at *1 (N.D. Ind. Aug. 21, 2013). Judge Miller denied plaintiffs request, since such information was not discoverable under Rule 26(b)(1). Id. at *2. Here, Biomet had already stated that these documents were included among the entirety of the document production, and the plaintiffs Steering Committee wants to know, not whether a document exists or where it is, but rather how Biomet used certain documents before disclosing them. Id. Nonetheless, Judge Miller encouraged Biomet to cooperate with this request, on the grounds that an unexplained lack of cooperation in discovery can lead a court to question why the uncooperative party is hiding something, and such questions can affect the exercise of discretion. Id. The outcome in Biomet might be different had the parties discussed predictive coding at the outset, instead of waiting until after the defendant had produced 2.5 million documents, and then demanding that the defendant switch its electronic discovery methods mid-stream. 6

8 1185 Predictive coding is the next wave in electronic discovery in employment litigation. It remains to be seen whether it will have staying power, or turn out to be a fad that gets replaced in a few years with some other technology. See Nair & Tookey, supra, National Law Journal, Jan. 28, 2013, at 11 ( So far, no case has taken predictive coding from soup to nuts from a discovery ruling through to the conclusion of discovery. ). But, employment lawyers need to familiarize themselves with its advantages and shortcomings, so that they can assess whether predictive coding is appropriate for electronic discovery in their case. And, if the parties agree to use predictive coding (or the court orders that it be used), then employment lawyers need to address the adequacy of each stage in the process of predictive coding, in order to ensure that the right documents get produced in a timely manner. II. Attorney-Client Privilege and Electronic Discovery. Regardless of how the responsive and relevant electronic documents are selected from the universe of electronic documents and s, the next step is to review them for attorney-client privilege and any other applicable privileges. Keyword searching may be one starting point in conducting a privilege review. For example, one can search for s containing the names of in-house or outside attorneys who may have provided legal advice or were otherwise involved in the adverse employment actions. As a threshold matter, not all communications with an attorney are necessarily privileged, since the privilege does not cover business advice, as opposed to legal advice. Neuder v. Batelle Pacific Northwest Nat. Lab., 194 F.R.D. 289, 293 (D.D.C. 2000) (discussion with in-house attorneys at meeting to discuss plaintiff s proposed termination would not be privileged if a business purpose took precedence over legal purpose). However, keyword searching is probably not an adequate substitute for an attorney s privilege review of each responsive document. Magistrate Judge Grimm, in a dispute involving violations of the copyright and patent statutes, and unfair competition claims, recently discussed the defendant s use of keyword searches which were inadequate, since the searches led to the production (instead of withholding) of 165 privileged documents. Victor Stanley, Inc. v. Creative Pipe, Inc., 250 F.R.D. 251 (D. Md. 2008). Although there was some dispute as to how the defendant actually conducted the keyword searches for privileged documents, and whether the non-searchable (scanned documents) were manually searched for privilege, id. at , there was no dispute that the plaintiff received 165 privileged documents. Magistrate Judge Grimm found that regardless of the keyword methodology used, there are well-known limitations and risks associated with them, and proper selection and implementation obviously involves technical, if not scientific knowledge. Id. at 260. Therefore, given the critical need to protect privilege, extra effort is required by the 7

9 1186 producing party: Use of search and information retrieval methodology, for the purpose of identifying and withholding privileged or work-product protected information from production, requires the utmost care in selecting methodology that is appropriate for the task because the consequence of failing to do so, as in this case, may be the disclosure of privileged/protected information to an adverse party, resulting in a determination by the court that the privilege/protection has been waived. Selection of the appropriate search and information retrieval technique requires careful advance planning by persons qualified to design effective search methodology. The implementation of the methodology selected should be tested for quality assurance; and the party selecting the methodology must be prepared to explain the rationale for the method chosen to the court, demonstrate that it is appropriate for the task, and show that it was properly implemented. Victor Stanley, 250 F.R.D. at 262 (emphasis added). Here, the Defendants have failed to demonstrate that the keyword search they performed on the text-searchable ESI was reasonable, and their attempts to argue that the volume of ESI needing review and time constraints presented them with no other choice is simply unpersuasive. Id. Moreover, this is not an instance in which a party inadvertently produced privileged information to an adversary, discovered the disclosure promptly, and then took immediate steps to inform the adversary that they had received the information inadvertently, thus demanding that it be returned. Id. at 263. While it is true that every waiver of the attorney-client privilege produces unfortunate consequences for the party that disclosed the information, the only injustice in this matter is that done by the Defendants to themselves. Id. Almost four months after the Victor Stanley decision, Congress approved a new evidentiary rule governing the inadvertent disclosure of privileged information in discovery. See Rule 502, Fed. R. Evid. (enacted Sept. 19, 2008). Rule 502(b), which governs inadvertent disclosure, codifies the common-law on waiver of the privilege through inadvertent disclosure: 502(b) Inadvertent disclosure. When made in a federal proceeding or to a federal office or agency, the disclosure does not operate as a waiver in a federal or state proceeding if: (1) the disclosure is inadvertent; (2) the holder of the privilege or protection took reasonable steps to prevent disclosure; and (3) the holder promptly took reasonable steps to rectify the error, including (if applicable) following Federal Rule of Civil Procedure 26(b)(5)(B). 8

10 1187 The cross-reference to Rule 26(b)(5)(B), Fed. R. Civ. P., is to the new clawback provision of the civil discovery rules: If information produced in discovery is subject to a claim of privilege or of protection as trial-preparation material, the party making the claim may notify any party that received the information of the claim and the basis for it. After being notified, a party must promptly return, sequester, or destroy the specified information and any copies it has; must not use or disclose the information until the claim is resolved; must take reasonable steps to retrieve the information if the party disclosed it before being notified; and may promptly present the information to the court under seal for a determination of the claim. The producing party must preserve the information until the claim is resolved. Rule 26(b)(5)(B), Fed. R. Civ. P. Rule 502(b) probably would not have changed the outcome in Victor Stanley, since it is clear from the facts and procedural history of that case that the second and third factors of the three-part test were not satisfied, i.e., defendants did not use adequate keyword search methods for doing a privilege review, and defendants failed to take prompt action after plaintiffs counsel notified them of the repeated disclosures of privileged materials. Other courts, in applying Rule 502, have similarly looked to whether the producing party took adequate measures, both in reviewing the documents, and in taking action after being notified of the release of privileged information. For example Judge Kollar-Kotelly, in a whistleblower retaliation claim, found that the defendant District of Columbia failed to take adequate measures at either stage, resulting in the waiver of privilege as to an with the agency counsel discussing certain matters relating to William s proposed termination. Although the District of Columbia requested return of that document at some unspecified time after it was produced, the plaintiff did not respond, and it is similarly undisputed that the District never followed up on its letter. Instead, the District of Columbia waited for two years and eight months, before taking any further action after the plaintiff did not return the privileged document. Williams v. District of Columbia, 806 F. Supp. 2d 44, 47 (D.D.C. 2011). Judge Kollar-Kotelly applied Rule 502 to find waiver of the privilege as to this document. Id. at The District failed to explain how it reviewed the documents for privilege, except in conclusory terms, and failed to explain why it took no action for nearly three years. This sort of indifference is fundamentally at odds with the principle that the attorney-client privilege must be jealously guarded by the holder of the privilege lest it be waived. Id. at 52 (quoting In re Sealed Case, 877 F.2d 976, 980 (D.C. Cir. 1989)). Therefore, the safe harbor in Rule 502(b) did not protect this privileged document. 9

11 1188 In contrast, Magistrate Judge Dolinger, in a patent infringement case, found that the plaintiff could invoke Rule 502(b), since the plaintiff took proper steps to rectify the inadvertent disclosure of privileged materials. Briese Lichttechnik Vertriebs GmbH v. Langton, 272 F.R.D. 369 (S.D.N.Y. 2011). Here, the plaintiff had produced 118 pages of documents (which was part of a larger, rolling production of over 8,000 documents), and only two weeks later, discovered that 14 of those pages were privileged, and sent a letter that same day requesting the immediate return or destruction of those 14 pages. Id. at 371. Defendants, however, took no action in response to this request, and instead filed a discovery motion seeking production of all other privileged documents on the grounds of waiver, and later claimed that the 14 pages were disclosed for tactical reasons. Id. at Magistrate Judge Dolinger rejected the tactical argument, and found that the plaintiff had acted promptly in demanding return, so that the Rule 502(b) safe harbor applied. It should be noted that this decision did not discuss the second factor, i.e., whether the privilege review of the documents prior to the production was adequate. Rule 502(d) allows the court to issue a discovery protective order that would preclude any finding of waiver: A federal court may order that the privilege or protection is not waived by disclosure connected with the litigation pending before the court in which event the disclosure is also not a waiver in any other federal or state proceeding. In other words, Rule 502(d) permits a producing party to open its files to an adversary, allows the adversary to decide what information it considers relevant and wants to use, and permits the disclosing party to assert a privilege thereafter. See G. P. Joseph, Impact of Rule 502(d), National Law Journal, Nov. 17, 2008, at 13. Thus, some courts are starting to encourage parties to consider requesting a Rule 502(d) provision as part of the Rule 26 order governing discovery. Doe v. Nebraska, 788 F. Supp. 2d 975, 986 (D. Neb. 2011). Other courts are unilaterally entering such orders over the objection of one or both parties, as a way to short-circuit discovery disputes. S2 Automation LLC v. Micron Technology, Inc., 2012 WL (D.N.M. July 23, 2012). One court applied a Rule 502(d) order to block the attempt by one party to retain and use privileged documents disclosed by the other party. Rajala v. McGuire Woods, LLP, 2013 WL (D. Kan. Jan. 3, 2013). In Rajala, a securities action, Magistrate Judge Waxse applied the Rule 502(d) order to reject defendant s attempt to retain a privileged document or otherwise to find waiver of the privilege as to the subject matter of that document. It is important to ensure that if there is a Rule 502(d) order, that it be carefully drafted in order to take into account the most dangerous threat posed by Rule 502(d) in the context of intentional disclosure of protected information namely that a party could use the privilege as both a sword and a shield through selective disclosure of privileged documents that are exculpatory. See Joseph, supra (suggesting model language for a Rule 502(d) order). 10

12 1189 The courts are divided on whether a clawback order must encompass each provision of Rule 502(b) in order to supplant the default test under that Rule. Some courts have held that unless all three prongs of Rule 502(b) (i.e., inadvertent disclosure; holder took reasonable steps to prevent disclosure; and holder promptly took reasonable steps to rectify the error in disclosing) are addressed in a clawback order, then the court will default to Rule 502(b) to fill in the gaps in controlling law. U.S. Home Corp. v. Settlers Crossing, LLC, No. DKC , 2012 WL , at *5 (D. Md. July 23, 2012) (collecting cases), reconsideration denied, 2012 WL (D. Md. Oct. 18, 2012). Other courts have held that a clawback order does not have to include specific detail under Rule 502(b) in order to supplant Rule 502(b), and have rejected attempts by the receiving party to argue that Rule 502(b) s provisions must be incorporated into a clawback order. Great-West Life & Annuity Ins. Co. v. American Economy Ins. Co., No. 2:11-cv-02082, 2013 WL , at *12-13 (D. Nev. Sept. 23, 2013), stay denied, 2013 WL (D. Nev. Nov. 6, 2013). That said, the better practice would be to ensure that a clawback order does address each of the three prongs of Rule 502(b), so that there is no dispute as to the law governing the application of the clawback order. Also, inadvertent disclosures of privileged documents to an expert witness may fall outside the scope of Rule 502. In Great-West Life, the defendant provided its expert witness with a number of documents to prepare his report, and later turned over those documents to the plaintiff, pursuant to the Rule 26(a)(2)(B) disclosure of information relied upon by the expert. Great-West Life, 2013 WL , at *15. Subsequently, the defendant realized that some of those documents in the expert witness production were, in fact, privileged documents, and demanded that the plaintiff return them. Magistrate Judge Hoffmann granted plaintiff s motion for a ruling that no valid privilege applies to these documents, since the privilege was waived through their disclosure under Rule 26, given the bright line rule that all information provided to a testifying expert is generally discoverable, other than drafts of the expert reports and communications with the party s attorney. Id. at *16. Hence, Advisory Committee Notes to Rule 26 explain that the facts or data considered by the witness in forming the opinions must be interpreted broadly to require disclosure of material considered by the expert, from whatever source, that contains factual ingredients. Id. (quoting Advisory Committee Notes to 2010 Amendments to Rule 26(a)(2)(B)). Thus, Rule 502 s clawback provision does not apply to documents that were provided to an expert witness, particularly where the expert stated in his report or deposition that he reviewed all the documents provided to him by the party that retained him. Id. at * Thus, at least one court has found that disclosures of privileged information to an expert witness can operate as a waiver that cannot be overcome by Rule 502. III. Ethical Issues Arising from an Employee s Use of Workplace Computers and for Privileged Communications with Outside Attorneys. 11

13 1190 A common situation in employment litigation arises when a current employee contacts an attorney in order to discuss workplace situations that may be discriminatory or otherwise violate the laws governing the employment relationship. In an ideal world, a current employee would only communicate with an attorney through a personal account, on a personal computer, and not by using a workplace computer or a workplace account. In reality, some employees use their workplace systems these communications, at least the initial stages. Then, when the employer s counsel reviews its electronic files, whether in investigating the employee s allegations or in responding to discovery requests, the employer uncovers those privileged s between the employee and her attorney. Has there been a waiver of the privilege? And, can the employer s counsel ethically make use of those communications? In August 2011, the American Bar Association issued two Formal Ethics Opinions on these issues. The first held, as a general principle, that an employee s use of workplace and computers for communications with a private attorney could waive the attorney-client privilege in those communications, and that the attorney should instruct her client not to use workplace and computers for these communications. See ABA Formal Op , Duty to Protect the Confidentiality of Communications with One s Client (2011). The second opinion held that when an employer s attorney discovered such s on the employer s server, the best practice was for the attorney to notify the employee s attorney and refrain from reviewing the s, but the employer s attorney was not obligated to do so. See ABA Formal Op , Duty when Lawyer Receives Copies of a Third Party s Communications with Counsel (2011). Over the past decade, courts have struggled with the issues that arise when an employee uses his workplace computer or workplace in order to communicate with a private attorney about potential or actual legal claims. Although the best practice is for attorneys to advise their clients that they should only conduct such communications on their home computer or personally-owned handheld device, and only through a personal account, some employees use workplace computers and for communications with their private attorneys. If the employee then brings legal claims against the employer, the employer s attorneys may well uncover these communications while reviewing the employer s accounts in order to find all s that are relevant to the legal claims, as part of routine monitoring of employees, or to discover misconduct that justifies termination (i.e., the after-acquired evidence doctrine). Some courts that have addressed this issue have held that any attorney-client privilege was waived by the employee through the use of the employer s computers for these otherwise privileged communications, while other courts have found that there was no waiver, such as where the employer authorized the employee s use of the workplace computer for purposes of participating in an investigation, or where the employee used a password to access his personal . 12

14 1191 In Stengart v. Loving Care Agency, Inc., 990 A.2d 650 (N.J. 2010), the New Jersey Supreme Court held that an employee had a common law reasonable expectation that the s she exchanged with her attorney using her personal, password-protected web-based account, but which she sent using her employer s computer, would remain private. The employee s expectation was reasonable because of the ambiguous language of the employer s electronic communications policy and the attorney-client nature of the communications. Id. at 663. The electronic communications policy acknowledged that occasional personal use of the employer s was permitted, did not address the use of personal, web-based accounts, and did not warn employees that the contents of s sent via personal accounts could be forensically retrieved and read by the employer. Id. Further, the s contained no illegal or inappropriate material that might have harmed the employer in some way if stored on its equipment. Id. The Stengart court warned employers, however, that even a carefully-worded electronic communications policy that banned all personal computer use and provided unambiguous notice that the employer could retrieve and read an employee s attorney-client communications, if accessed on a personal, password-protected account using the employer s computer system likely would not be enforceable. Id. at 665. However, Stengart is a state court opinion that has not been followed by all other courts in addressing this issue. Thus, attorneys for employees need to require that their clients use only personal computers and to communicate with the attorney. IV. Ethical Issues Arising from Attorneys Accessing an Employee s Postings on Social Media and Social Networking Sites. Social media, for better or worse, is prevalent in our society. Over the past several years, there has been extensive case law as to whether an employee s postings on social media may be protected activity under the anti-retaliation statutes or the National Labor Relations Act. There has similarly been extensive case law as to whether a party can be required to produce, in discovery, her social media pages or postings. The courts have reached divergent results, particularly depending on whether the information sought was from the public postings or the private sections of the employee s social media (e.g., the public and private sections of a Facebook page). This section addresses the ethical issues of whether an employer s attorney may access an employee s social media information in order to obtain information to defend the employer against the employee s claims. Several bar ethics opinions provide useful guidance on this point. The Philadelphia Bar Association issued an opinion in March 2009, in response to a query as to whether an attorney could access the Facebook and MySpace accounts of a witness who was deposed and provided testimony helpful to the other party, where that witness was herself not represented. See Philadelphia Bar Association, Professional Guidance Committee, 13

15 1192 Opinion (March 2009). The attorney thought that these sites could have useful impeachment information, but when he tried to access those sites, he determined that her accounts were blocked, and that access could only be obtained with the witness s permission. Hence, the attorney was going to ask a third person to go to those sites, contact the witness and seek to friend her, to obtain access to the information on the pages.... [and] would not reveal that he or she is affiliated with the lawyer or the true purpose for which he or she is seeking access, namely to provide the information posted on the pages to a lawyer for possible use antagonistic to the witness. Id. at 1. The proposed conduct would be covered by Rule 5.3 of the Rules of Professional Conduct, Responsibilities Regarding Nonlawyer Assistants, since lawyers are responsible for the conduct of such assistants. Id. at 2. And, the proposed conduct would violate Rule 8.4(c), Misconduct, which prohibits lawyers from engag[ing] in conduct involving dishonesty, fraud, deceit or misrepresentation. Id. at 3. The third person s contacting the witness to request access to her Facebook and MySpace pages would be deceptive since the third person purposefully concealed the purpose of obtaining that access. Id. Moreover, the proposed conduct would also violate Rule 4.1, Truthfulness in Statements to Others. Id. at 4. The New York City Bar Association reached the same result, under the similar New York Rules of Professional Conduct. See N.Y. City Bar Association, Formal Opinion , Obtaining Evidence from Social Networking Websites (2010). Although both the New York City Bar Association and the Philadelphia Bar Association are voluntary bar associations, their ethics opinions may be influential both in those jurisdictions and elsewhere, and should be considered before attempting to use any deceptive means in social media discovery. At the same time, there is no barrier to an attorney accessing the public pages of an adversary s social media websites, on the grounds that the adversary has made that information freely available to the general public. See N.Y. State Bar Association, Opinion 843 (Sept. 10, 2010). The New York State Bar Association concluded that: A lawyer who represents a client in a pending litigation, and who has access to the Facebook or MySpace network used by another party in litigation, may access and review the public social network pages of that party to search for potential impeachment material. As long as the lawyer does not friend the other party or direct a third person to do so, accessing the social network pages of the party will not violate Rule 8.4 (prohibiting deceptive or misleading conduct), Rule 4.1 (prohibiting false statements of fact or law), or Rule 5.3(b)(1) (imposing responsibility on lawyers for unethical conduct by nonlawyers acting at their direction). 14

16 1193 Id. at 6. The New York State Bar Association also noted that attorneys should not attempt to friend a represented party in order to access the private sections of their social media sites, and should exercise caution before friending unrepresented parties: Id. at n.1. If a lawyer attempts to friend a represented party in a pending litigation, then the lawyer s conduct is governed by Rule 4.2 (the no-contact rule), which prohibits a lawyer from communicating with the represented party about the subject of the representation absent prior consent from the represented party s lawyer. If the lawyer attempts to friend an unrepresented party, then the lawyer s conduct is governed by Rule 4.3, which prohibits a lawyer from stating or implying that he or she is disinterested, requires the lawyer to correct any misunderstanding as to the lawyer s role, and prohibits the lawyer from giving legal advice other than the advice to secure counsel if the other party s interests are likely to conflict with those of the lawyer s client. 15