The attorney-client privilege is a long-standing, well-established doctrine that has seen its boundaries shift over time. But over the decades and in different jurisdictions, the core purpose and the essential elements of the privilege doctrine remain largely unchanged. This highlights its significance to the attorney-client relationship in American jurisprudence. Yet the privilege doctrine is not static. It has evolved as legislation and case law have defined the nature and scope of the doctrine: Was there an existing attorney-client relationship? What about a communication? Was it made in confidence? Was legal advice requested or rendered? Was the privilege waived? Do any exceptions apply? These questions arise constantly.

Historically, privilege analysis focused on the application of a test-based approach to the doctrine. This can create some inherent difficulties in applying the doctrine to a discrete set of facts. For example, almost forty years ago, writing for the majority in Upjohn Co. v. United States, Justice William Rehnquist of the U.S. Supreme Court, prefaced the court's opinion by remarking: "the parties and various amici have described our task as one of choosing between two 'tests' … We are acutely aware, however, that we sit to decide concrete cases and not abstract propositions of law. We decline to lay down a broad rule or series of rules to govern all conceivable future questions in this area, even were we able to do so." 449 U.S. 383, 386 (1981).

Justice Rehnquist's comment that Upjohn does not "lay down a broad rule or series of rules to govern all conceivable future questions" regarding the scope of attorney-client privilege seems prescient in light of modern digital discovery practices. The post-Upjohn litigation landscape is riddled with parties involved in lengthy, voluminous discovery disputes over primarily electronic communications and documents. But as Justice Rehnquist observed later in Upjohn, "if the purpose of the attorney-client privilege is to be served, the attorney and client must be able to predict with some degree of certainty whether particular discussions will be protected." Id. at 393. Balancing the need for predictable application of the privilege with Justice Rehnquist's acknowledgment that courts "sit to decide concrete cases" is a harder task than ever, given the sheer amount and complexity of disputed communications modern courts are tasked with evaluating.

Thus, in an era of discovery dominated by email, text messages and the like, what balance (if any) has been struck between governing privilege standards and case-by-case analysis of disputed communications? Below are some thematic observations based on selected 2019 cases regarding the scope of attorney-client privilege as applied to digital communications.

Redaction Standards and Practices

Presumably, a corollary of the desirability of predictable application of attorney-client privilege is the need for reliable standards of redaction when courts decide that only portions of emails or documents may be withheld based on privilege. But a review of the selected 2019 cases shows that it is difficult to point to a consistent pattern in the courts' redaction practices for two main reasons.

First, courts generally refrain from quoting freely from such communications in published opinions, given confidentiality considerations. Instead, courts often describe the communications at issue and attempt to delineate the attributes of the communications that make them privileged or not. Sometimes, this approach suffices to illustrate which features of the communication invoke privileged status. For example: "[t]he first redaction in [defendant's employee's] email … discusses legal interpretations of the license agreements between [plaintiff] and [defendant] and specifically references [defendant in-house counsel]." Dolby Labs. Licensing v. Adobe, No. 18CV01553YGRDMR, 2019 WL 4082784, at *9 (N.D. Cal. Aug. 29, 2019).

Second, courts must routinely pore over virtual mountains of email, text messages and other electronic communications. Given that privilege analysis is a document-by-document endeavor, it is not surprising that courts' descriptions of disputed communications are sometimes little more than bare-bones recitations of the applicable privilege standard. A recent California decision is particularly forthright: "[t]o address each redaction separately in this order is simply not practical and would waste scarce judicial resources." Bruno v. Equifax Info. Servs., No. 2:17-CV-327-WBS-EFB, 2019 WL 633454, at *9 (E.D. Cal. Feb. 14, 2019).

Interestingly, in Pearlstein v. BlackBerry Limited, the sheer number of communications at issue resulted in inconsistent treatment of similarly situated communications. For instance, it is not clear from the opinion why "[r]edaction is neither feasible nor appropriate" for email chain No. 245 while email chain Nos. 289 and 298 may be redacted "to the extent [such communications] contain any [legal] analysis." No. 13CV07060CMKHP, 2019 WL 1259382, at *12, 15, 18 (S.D.N.Y. March 19, 2019). Each aforementioned email chain was between defendant's in-house counsel and defendant's non-legal personnel, and contained a discussion of the same document. Pearlstein at *20. Perhaps the privileged communications in email chain No. 245 were so inextricably intertwined with the non-privileged portions that redaction was not feasible. But when certain other courts redact at the level of each individual word (see AdTrader v. Google, No. 17CV07082BLFVKD, 2019 WL 4221586, at *3 (N.D. Cal. Sept. 5, 2019) ("Only the first four words of the sentence are privileged.")), it is perplexing why some privileged communications are considered to be intrinsically entangled with non-privileged information while other privileged communications may be redacted with a scalpel.

Unique Features of Digital Communications

Digital communication technology is familiar to most attorneys at this point, but the legal ramifications of such technology's ubiquity are far from settled. With respect to the attorney-client privilege, recent cases feature growing consideration of the distinctive characteristics of digital communications such as the "cc" function of email and the timestamps on text messages. For example, in Bruno, the court noted that merely "cc"-ing an attorney on an email is not sufficient to invoke attorney-client privilege, and cited precedent to the same effect. Bruno at *5. In Dolby, the court divided a text message exchange into privileged and non-privileged portions based on time-stamp designations. Dolby at *11.

In addition, in Bruno, the court held that for purposes of privilege analysis, it must consider email attachments as separate communications from the emails themselves, citing supporting California federal precedent. Accordingly, the defendant in that case was compelled to produce the attachments. Moreover, the Bruno court held that a draft email was privileged to the same extent as the final version of the email. Bruno at *8.

Ease of Waiver of Privilege

The myriad ways in which privilege can be waived or otherwise compromised are not new to the digital era, but the prevalence of digital communication gives attorneys and their clients countless opportunities to make a mistake. Consider the Bruno holding above. If an email attachment is a separate communication, what about the draft email? How many changes in a draft email transform it from version one to version two, which version is also potentially a separate communication?

Complicating matters further, the "cc" function of email and similar group communication technology makes it absurdly easy to waive the privilege via third-party disclosure. After all, many lawyers may be guilty of judicious use of the "Reply All" function. Even for the most careful practitioners, there are simply too many chances to make an error. Imagine a hypothetical attorney who, with 99% accuracy, is able to discern whether communications (s)he engages in are privileged or not. Given 100 chances, the probability the attorney makes no mistakes is about 37%. With 1,000 chances, it is about .004%. Of course, a solitary mistake may not necessarily be fatal to preserve privilege, but depending on the applicable jurisdiction, doctrines such as "subject matter waiver" and the "shield and sword" principle may imperil multiple related communications based on a single, unintentional disclosure.

The selected 2019 cases illustrate the variety of potential privilege pitfalls that could be lurking. For example, in Curtis v. Progressive Northern Insurance Company, defendant waived privilege with respect to several email chains by including a third-party vendor in the email chains. Defendant unsuccessfully argued that the vendor was the functional equivalent of defendant's employee and therefore protected under defendant's privilege. No. CIV-17-1076-PRW, 2019 WL 1937596, at *5-6 (W.D. Okla. May 1, 2019). In Bruno, defendant's unsuccessful invocation of the "common interest" exception to waiver resulted in the production of not only the disputed communications, but also "all communications related to [the disputed communications]" under the "subject matter waiver" doctrine. Bruno at *6, 10. Finally, in Kleeberg v. Eber, defendant's disclosure to his daughter (and business partner) of an email containing legal advice resulted in waiver of the privilege because the legal advice was rendered to defendant in his individual capacity, rather than in his capacity as an officer of his business. No. 16CV9517LAKKHP, 2019 WL 2085412, at *23 (S.D.N.Y. May 13, 2019).

As the privilege doctrine is analyzed in future discovery disputes, Justice Rehnquist's observations in Upjohn should be remembered, especially as technology's role in privilege analysis continues to grow.

Devika Kewalramani is a partner at Moses & Singer and chair of its legal ethics and law firm practice. She is a former co-chair of the Council on the Profession of the New York City Bar. Peter Kimble is an associate at the firm. Associate Matthew Handler assisted with the preparation of this article.