Hard to believe, but it took until August 29, 2019 for the North Carolina Court of Appeals to tackle the issue of eDiscovery for the very first time. But tackle it the court did.
In the case of Crosmun v. Trustees of Fayetteville Technical Community College, the court was presented with its “first opportunity to address the contours of eDiscovery within the context of North Carolina common and statutory law regarding the attorney-client privilege and work-product doctrine.” Former community college employees sued the school alleging retaliatory dismissal in violation of the North Carolina Whistleblower Act. The plaintiffs propounded routine discovery on the defendants, including for electronically stored information (ESI) retained in the school’s computers and servers.
The trial court ultimately entered an order compelling discovery. In connection with that order, it also entered an order, as requested by the plaintiffs, providing for a forensic examination of the defendants’ computer files. The trial court adopted certain protocols sought by the plaintiffs, which provided that a computer forensic expert retained by the plaintiffs would conduct a forensic examination of the defendants’ computer files, conducting keyword searches for all responsive data, as well as keyword searches for potentially privileged data. Documents retrieved that were not identified from keyword searches as privileged were to be delivered directly to the plaintiffs, without the defendants being able to review them for privilege. The defendants appealed from this order, contending that it amounted to an involuntary waiver of their attorney-client privilege and the work-product doctrine.
Before addressing the precise issue before the court, the comprehensive opinion, written by Judge Lucy Inman, first analyzed the broader contours of eDiscovery, observing that “ESI has become so pervasive that the volume of ESI involved in most cases dwarfs the volume of any paper records.” Despite that, the court noted, “North Carolina authority regarding eDiscovery is bare bones.” More to the point, “[n]o statute, procedural rule, or decision by this Court or the North Carolina Supreme Court has delineated the parameters of eDiscovery protocols with respect to the protection of documents and information privileged or otherwise immune from discovery.” The court was therefore left to gather, digest, and analyze decisions from other courts around the nation, as well as principles articulated by the Sedona Conference, on which the appellate panel relied extensively.
The court stated that forensic examinations of ESI “may be warranted when there exists some factual basis to conclude that the responding party has not met its duties in the production of discoverable information.” But even when such an examination is proper, “any protocol entered must take into account privileges from production that have not been waived or otherwise lost.” Courts ordering forensic examinations must protect against disclosure of: a) trade secrets; b) private and confidential information, such as personnel evaluations and payroll information; and c) attorney-client and work-product communications. In addition, courts must be mindful of computer examinations which would: a) unreasonably disrupt ongoing business; b) endanger the stability of computer systems and files; or c) place the responding party’s computer systems at risk of a data security breach.
The Sedona Principles provide recommendations to ensure these interests are protected. They include documenting an agreed-upon examination protocol, recognizing the rights of non-parties, such as employees, patients, and other entities, and narrowly restricting the search to protect confidential and personal information and to avoid access being granted to information unrelated to the litigation. Invariably, courts applying these principles have permitted the producing party to conduct a prior privilege review of all ESI being produced.
Turning to the facts before it, the court found two reversible errors in the trial court’s order for a forensic examination of the defendants’ computer systems. First, it permitted the plaintiffs’ expert, rather than an independent third party, direct access to the entirety of defendants’ ESI “absent regard for Defendants’ privilege.” Second, it ordered the delivery of responsive documents to the plaintiffs prior to permitting the defendants to review them for privilege. “In both instances, the protocol compels an involuntary waiver, i.e., a violation of Defendants’ privileges.” It noted that at oral argument, the plaintiffs’ counsel conceded that the appointment of a court-appointed special master would have been a more neutral way to permit such a computer search.
Comparing this process to the analog world prior to the advent of eDiscovery, the court posited a plaintiffs’ expert being permitted to “photocopy all of Defendants’ documents (including those in their in-house counsel’s file cabinets)” and to review them “without Defendants having had an opportunity to conduct their own review of those copies first. Such a process would violate Defendants’ attorney-client privilege as a disclosure to the opposing party. . . . The digital equivalent does so as well.”
The plaintiffs argued that the protocol order already protected against the defendants’ privilege claims by requiring their expert to screen for privilege documents that hit upon keyword search words designed to ferret out those with potential privilege claims. The appellate court, however, was unconvinced: “While the use of search terms assists in the disclosure of privileged materials, it is far from a panacea. All keyword searches are not created equal; and there is a growing body of literature that highlights the risks associated with conducting an unreliable or inadequate keyword search or relying exclusively on such searches for privilege review.” The selection and implementation of keyword searches is fraught with “well-known limitations and risks.” That is why quality-control procedures are so vital prior to the production of any responsive information to an opposing party.
The court held that this problem is not ameliorated by a clawback provision, for the following reasons: First, the plaintiffs’ agent is granted direct access to the defendants’ privileged information. Second, that agent must then produce any ESI not flagged by the privileged keyword searches to the plaintiffs without any opportunity for privilege review by the defendants. Third, the defendants must then attempt to clawback that information, “reducing their privilege to a post-disclosure attempt at unringing the eDiscovery bell. Such compelled disclosure of privileged information is contrary to our law concerning both attorney-client privilege and work-product immunity.” The court therefore vacated the trial court’s order on the forensic examination of the defendants’ computer systems.
The Court of Appeals provided guidance to the trial court to fix these problems the second time around, guidance that trial judges throughout North Carolina are likely to follow when facing similar, vexing, eDiscovery disputes:
First, it is advisable for the court to appoint a special master or independent expert to perform the forensic examination as an officer of the court. “Such an appointment appears to be the commonly accepted approach in other jurisdictions and is consistent with the leading treatises on eDiscovery. That removes the concern that privileged information will be disclosed to an agent of one party or the other.”
Second, it is advisable for the responding party to have some opportunity to review the keyword search its prior to production of responsive ESI to the opposing party. Further, the trial court can order that any ESI produced under the protocol is confidential within the meaning of the applicable protective order and is subject to clawback without waiver of any privilege or work-product immunity.
But whatever the trial court finds most appropriate on remand, the protocol adopted must not “deprive the Defendants of an opportunity to review responsive documents and assert any applicable privilege[.]”
Crosmun is a well-reasoned, well-written opinion that is likely to be North Carolina’s seminal case on eDiscovery for years to come. It should be reassuring to the corporate world, which should no longer fear voluminous, unrestricted forays into companies’ computer systems without careful forethought and protection from North Carolina’s judges. A good opinion indeed.