This article is the fifth installment in a series on e-discovery issues and areas that offer inside counsel the greatest opportunities to reduce risks and save costs. Read parts one, two, three, and four.


Organizations in litigation spend prodigious amounts on document review, in large part to control the significant risk that a privileged document will be inadvertently produced.

In 2003, in the second of her now-famous e-discovery opinions in the Zubulake v. UBS Warberg case, Judge Shira Schindlein of the Southern District of New York wrote that the cost of reviewing documents could, in part, be controlled by the producing party because that party could enter into a clawback agreement that would allow it to “undo” a document production and demand the return of any inadvertently produced documents. As the judge said, “many parties to document-intensive litigation enter into so-called ‘clawback’ agreements that allow the parties to forego privilege review altogether in favor of an agreement to return inadvertently produced privileged documents.”

There were, however, two problems with this suggestion.

  1. Because she was discussing clawback “agreements,” it was not altogether correct to suggest that the producing party had any degree of control, at least insofar as the other side might not agree.
  2. Even assuming the requesting party was so agreeable, a significant risk still persisted that a non-party—not bound by the agreement—could come to claim that the inadvertent production nonetheless constituted a waiver as to that non-party, such that whatever the parties’ agreement between themselves, the non-party would be entitled to the otherwise privileged document and perhaps even all other privileged documents relating to the same subject matter.

But in 2008, Congress enacted Federal Rule of Evidence 502, making a number of substantive changes to the rules of privilege. Among them, Rule 502 adopted a uniform standard for inadvertent production by finding that in general, courts would not find a waiver so long as the producing party had taken reasonable precautions to prevent the disclosure and had promptly taken appropriate steps to correct the error.

Further, the rule expressly allowed parties to enter into the well-established clawback agreements regarding the effect of disclosures both inadvertent and intentional of privileged material, and that those agreements would be effective not only between the parties, but also for non-parties if the agreement is memorialized into a court protective order. Thus, clawback agreements can now be entirely effective on non-parties, as long as they are entered as a court order. Moreover, this can hold true irrespective of the efforts made by the parties to prevent disclosure or correct error.

Regrettably, too few parties take advantage of these protections, and for an organization that is party to a federal litigation involving a substantial amount of electronically stored information (ESI), there is simply no reason not to enter into such a clawback agreement. In certain litigations, as suggested by Judge Schindlein, this may allow a party to forego the entire expense of a privilege review by simply producing all relevant documents subject to the clawback. However, understanding that bells cannot be unrung, parties can nonetheless save tremendous cost through word searches of counsel names and terms like “attorney-client privilege,” or they might opt to still undertake a full privilege review, but now with the peace of mind that the clawback will operate as a mighty strong insurance policy.

Critically, however, when entering into a clawback agreement and proposed protective order, the parties should be sure to spell out the applicable privileges or protections that will not be waived and that efforts to protect against or correct waivers will not be considered in any waiver determination. If a clawback agreement memorialized into a protective order fails to be explicit on these points, a court might take a formalistic approach and look to these efforts in determining whether to permit a clawback. At least one court has already done so.

In a case involving the bilateral production of ESI, both parties will generally see the reasonableness of such an arrangement. But when one party has little or no data to produce, the opposing party may refuse to agree. In those cases, the party with the data can and generally should nonetheless unilaterally move the court to enter such an order.

There is nothing in Federal Rule of Evidence 502 requiring the parties to agree to the court order protecting against waiver, and Federal Rule of Civil Procedure 26(c)(1) provides that a court may, “for good cause,” enter a protective order to prevent “oppression, or undue burden or expense.” When such an order will allow a party to conduct and respond to discovery in an expeditious manner, without the need for time-consuming and costly privilege reviews—especially in cases involving e-discovery—good cause should be clear.

In short, clawbacks are cheap and strong insurance. Whenever your company is faced with a significant production of ESI, be sure one is in place as a protective order, either through an agreed or unilateral motion to court.