You’re beginning discovery on a case. When you meet with opposing counsel, you both agree that a formal 502(d) court order isn’t necessary here. Instead, you enter into a clawback agreement. Should anyone accidentally hand over privileged documents, both parties agree, the other side will return or destroy their copies and the inadvertent disclosure will not count as a waiver.
You should be good to go, right? Not exactly. A clawback agreement might not be enough to protect against a waiver of attorney-client privilege when a court deems the production of privileged docs to be “reckless,” as Irth Solutions, LLC v. Windstream Communications LLC, a case out of the Southern District of Ohio, demonstrates. And it’s not just discovery novices getting it wrong; this discovery debacle involves one of the nation’s most reputable law firms.
Irth, Windstream and Fire
Today’s lesson on privilege waiver comes out of dispute between a telecom utility and a company that makes software for owners of subsurface assets to manage so-called “dig notices,” essentially alerts that construction or excavation will be taking place where a utility line is located.
Irth, the software company, accuses Windstream Communications, the utility, of breach of contract and fraud stemming from a dispute over Windstream’s alleged refusal to pay for contract overages.
And that’s where the excavation really begins, for our purposes. At the outset of discovery, the parties—Windstream represented by attorneys from BakerHostetler, an AmLaw 100 firm, Irth by Walker Novack Legal Group—reached an agreement regarding the production of ESI. Central to the later dispute was the agreement’s clawback provision, which was memorialized thus:
The parties agreed that a formal court order under Fed. R. Evid. 502(d) was not necessary based on the scale of the case, but that the parties would agree among themselves as follows:
- If a producing party discovers that it has inadvertently produced a document that is privileged, the producing party will promptly notify the receiving party of the inadvertent production.
- The receiving party will promptly destroy or return all copies of the inadvertently-produced document.
- Inadvertent production of privileged documents does not operate as a waiver of that privilege.
That clawback provision did not provide the broad protection against privilege waivers that the parties may have wished, however, as Windstream would later find out.
Privilege Review Gone Wrong
Windstream’s privilege problems offer a comedy of eDiscovery errors. Windstream’s production came in two phases. In both productions, dozens of privileged documents, representing a shockingly high percentage of produced files, were turned over to the other side—that is, the same privileged documents were produced twice.
First, Windstream’s outside counsel, the attorneys with BakerHostetler, were apparently unaware of the company’s entire in-house team. When they conducted the initial privilege review, the outside attorneys explained, they overlooked one of the three Windstream in-house lawyers involved in the dispute.
Windstream also failed to produce a privilege log until several days after its initial production. And though it claimed it had withheld five to ten documents for privilege, the court subsequently found that only four documents had been withheld, totaling five pages.
An attorney for Windstream, when preparing a post-production privilege log, then realized that 43 privileged documents had been turned over to plaintiffs. By the time Irth’s attorneys were notified, 12 days after the production, most of the production had already been reviewed.
Irth’s counsel was unsympathetic to Windstream’s claims that the production was inadvertent, given the three months taken to turn over ESI and the lack of a privilege log. “[I]t is unbelievable that a firm with Baker Hostetler’s reputation would make such an inadvertent mistake,” the attorney responded—then refused to return the documents.
But wait, there’s more.
As the dispute over the privileged documents played out, Windstar was asked to produce the initial documents a second time, as some of the first production was not in a searchable form. This second production involved the same documents as the first and, despite the continuing fight over privileged, included the exact same privileged docs. “Thus, in the midst of arguing to this Court that it should protect Defendant’s attorney-client communications and award it fees and costs,” the court later wrote, “defense counsel again produced the privileged documents.”
Court "Unconvinced" of Meaningful Review
Windstream eventually moved for sanctions over Irth's refusal to destroy or turn over the privileged documents. But the court was far from sympathetic to Windstream's plight, largely rejecting the host of explanations for why Windstream's lawyers failed to properly catch and redact the privileged documents.
The documents, Windstream’s lawyers asserted, underwent two levels of review. First, a seventh-year BakerHostetler associate gave the documents an “initial review.” They were then passed on to a second-year BakerHostetler associate, who reviewed the documents on a legacy platform which “provided for coding for attorney/client privileged documents.” Finally, the docs were checked once again by the senior associate, then produced.
The second production, containing the same privileged files, was the result of a mistake on the part of the litigation support staff, Windstream’s lawyers claimed. The privileged docs had been redacted or removed from the set, but when it came time to produce, the support staff sent Irth the original files, rather than using the updated version.
U.S. Magistrate Judge Kimberly A. Jolson, however, seemed to question whether a review was conducted at all. The overlooked in-house counsel was not that hard to spot, even in a crowd of thousands of documents. Her email signature, after all, included “Counsel to Director of Government Contract Compliance.”
A staggering amount of Windstream’s improperly produced documents also included terms that shouldn't have been difficult to spot. As Judge Jolson writes:
Upon this Court’s request, defense counsel submitted the forty-three documents (146 pages) for in camera inspection prior to the hearing. A review of those documents revealed that fourteen of the forty-three, or more than 32%, contain the word “legal.” Stated differently, almost 1/3 of the documents contained the word “legal.” Of the documents containing the word “legal,” ten contained the question “Michelle, any guidance from legal . . . ?”
Now, there are certainly several explanations as to why such documents may have been overlooked. The technology used here, first conceived in 1982, is… not exactly cutting edge, and certainly not a model of intuitive, simple software—potentially exacerbating the risks of human error. Further, it’s not unheard of for miscommunications and misunderstandings to lead to the wrong documents being produced.
But, as the court noted, “the utilization of any basic key-word search would have flagged these documents for additional review.” Further, Windstream’s attorney’s equivocal language—the junior attorney “believes she performed the privilege review"—was “curious” at best.
“Indeed, the Court is unconvinced that any meaningful review of the documents occurred,” Judge Jolson wrote. She continued:
It is also worth noting that the first production was not the result of a technical error or third party-vendor mistake. Instead, attorneys reviewed a limited number of documents and made critical and reckless mistakes. This is not a case in which defense counsel was reviewing hundreds of thousands of documents and a few managed to evade review—a situation that is likely and almost unavoidable in massive document reviews. Instead, this is a case in which only 1400 readable pages were produced initially, and of those pages, 146 pages—or more than 10% of the entire production—were missed.
Three Approaches to Clawback Agreements and Rule 502(b)
At this point, the conclusion to this discovery drama shouldn’t be too surprising. Judge Jolson denied Windstream’s motion for sanctions and declined to order Irth to turn over the privileged documents. Given the recklessness of the production, attorney-client privilege had been waived, the judge concluded.
Under Federal Rule of Evidence 502(b), privileged documents produced during discovery may survive production without waiving privilege, but only if the production is inadvertent. Inadvertent disclosure, the court notes, even encompasses negligent disclosure and the court assumed, arguendo, that Windstream’s production was inadvertent here. Rule 502(b) further requires that, to preserve privilege despite an inadvertent disclosure, the disclosing party must have taken “reasonable steps to prevent disclosure” and “reasonable steps to rectify the error.”
But, Windstream argued, the parties’ claw back agreement should govern the privilege issue—and that agreement’s only trigger was an inadvertent production. Rule 502(e), Windstream noted, gives controlling effect to such agreements. Despite 502(e), the court was unconvinced that the clawback agreement would control:
Defendant fails to acknowledge, however, that despite Rule 502’s goal of creating uniformity, courts still dispute how to analyze inadvertent disclosures when a cursory clawback agreement exists and alleged carelessness caused an inadvertent production.
Indeed, the courts have developed three distinct approaches to the interplay between clawback agreements and Rule 502(b), which the court surveyed in turn.
Under the first approach, “courts have held that a clawback agreement (no matter how cursory) requires the return of inadvertently produced documents, regardless of the care taken by the producing party.” This approach has been used by district courts in Kansas, New Jersey, and Washington, D.C. However, the court rejected this approach, finding it “inconsistent with the underpinnings of Rule 502” and “an abdication of the Court’s role to interpret the parties’ agreement in this case.”
Under the second framework, protective orders with clawback provisions protect against waiver “unless the document production process itself was ‘completely reckless.’” Courts in the Second Circuit “have largely followed this approach,” the court notes, citing caselaw from both the Eastern and Southern Districts of New York. This approach, Judge Jolson explains, seeks to recognize the value of clawback agreements, without allowing them to “shield a party from the consequences of reckless conduct.”
Finally, under the third method, courts allow parties to contract around Rule 502(b)’s requirements only to the extent that the agreement provides “concrete directives regarding each prong of Rule 502(b).” That includes defining inadvertence, stating what precautionary measures are required, and establishing post-production responsibilities. In cases where such specificity isn’t present, “Rule 502(b) is interstitial, filling the silent gaps.”
Applying both the second and third approach, then, Windstream’s actions during discovery were insufficient to preserve privilege. Under approach number two, Windstream’s review and production was reckless, the court found, thus destroying the clawback agreement's protections. Applying the third approach, with the clawback agreement requiring nothing but inadvertence, Rule 502(b)’s reasonable steps requirements would still apply—and the reckless production here would not satisfy them.
The court did not seem to struggle in making its recklessness determination. There was the simple fact that more than 10 percent of all documents produced had been privileged, that those documents were full of easily discoverable terms, and that the inadvertent production was completed twice. This was a case where, the court explained, “normal cracks became chasms.”
This may not be the final word on the matter, however. Windstream is currently seeking to have the court order stayed and plans to file their objections to the order today, August 16th. In the meantime, this case is an important reminder that even a broad clawback agreement may not be enough to protect against a poorly done review.