Concept
Confidential Business Information
Protective orders and disclosure of protected material
Confidential Business Information
Rule 26(c)(1)(G) lets a court require that a trade secret or other confidential research, development, or commercial information not be revealed, or be revealed only in a specified way.1 The rule gives the court two settings, and the choice between them depends less on how secret the information is than on whether the case can be fairly decided without it. No true privilege protects a trade secret, so a business cannot refuse discovery by calling its files secret.2 What the business can ask for is control over the manner of disclosure.
The business interest at stake is concrete. In Chudasama v. Mazda Motor Corp., the Eleventh Circuit observed that the economic harm of letting commercial competitors see a manufacturer’s proprietary information may well exceed the manufacturer’s potential liability in the lawsuit.3 A court that orders such material produced therefore has reason to say exactly who may see it and what they may do with it.
This section first takes up the choice between withholding and conditional disclosure, and the conditions courts attach. It then turns to the two standing orders commercial litigants use to carry those conditions out: sharing orders, which settle who may receive protected material, and umbrella orders, which settle how material comes within the order.
Conditional Disclosure of Trade Secrets
Rule 26(c)(1)(G) authorizes an order that a trade secret or other confidential research, development, or commercial information not be revealed or be revealed only in a specified way.4 The clause names two outcomes but supplies no test for choosing between them. The surrounding doctrine supplies one. Because no true privilege protects trade secrets, disclosure will be ordered if the issues cannot be fairly adjudicated without the information. Even then the court must use its discretion to avoid unnecessary disclosure, and it may attach conditions that protect the party compelled to disclose.5
Privilege, where it applies, takes material out of discovery. A trade secret gets no such treatment, and that changes the question a court asks. The court asks whether the case needs the information and, if it does, how the other side will receive it. Each half of that inquiry answers a different interest. The necessity half keeps the label “trade secret” from becoming a way out of discovery, and the conditions half keeps compelled disclosure from exposing the owner more than the case requires.
The court has one more tool for the moment of production. It may order the parties to file specified documents or information simultaneously in sealed envelopes, to be opened as the court directs.6
The treatise’s own illustration shows the conditional setting at work. One party seeks another’s trade secrets, and the judge might allow the plaintiff to discover them while preventing the plaintiff from disclosing the information to any third party.7 The other setting follows from the same logic. When the requesting party does not need the information, the premise for compelled disclosure is missing, and the court may order that the information not be revealed at all.8
The two demands in that scene are equally sensitive, so sensitivity cannot be what separates them. Necessity decides whether the information comes out, and conditions decide how it comes out. The more the case depends on the information, the less room the owner has to keep it back, and the more of the court’s attention goes to the terms of disclosure.
The party claiming protection must also be able to say what it is protecting. A trade secret owner has to articulate, in concrete terms, exactly what it claims as its trade secrets. The claim need not be as detailed as a patent application, but it cannot be a vague or abstract concept.9 A party that invokes Rule 26(c)(1)(G) for “our proprietary know-how” and nothing more has described its interest without identifying a secret.
Gonzales v. Google shows what conditions look like when a court compels production of information a company calls a trade secret. The government sought material from Google’s proprietary search index, and Google had been unable to negotiate the terms of the protective order in the underlying litigation. The court recognized that selective disclosure of protectable trade secrets is not per se unreasonable and oppressive when appropriate protective measures are imposed. It also noted that the government had considerably narrowed its request, so the risk of trade secret disclosure was substantially mitigated.10 The court then ordered Google to confer with the government on a protocol for randomly selecting and promptly producing 50,000 URLs from its database, on three conditions. Google need not disclose proprietary information about its database in developing or implementing the protocol. The government pays Google’s reasonable cost of formulating and implementing the extraction. Whatever Google discloses falls under the protective order in the underlying case.11 The court denied the motion to the extent it sought users’ search queries.12 It also ordered the parties to show cause why a “Confidential” designation under the existing protective order would be insufficient protection, which treats that designation as the starting point and leaves room to argue for more.13
Each condition answers a distinct harm. A sample in place of the full index narrows what leaves the company. The protocol condition keeps the act of producing from exposing proprietary details. The cost condition places the expense of the extraction on the party who asked for it. The protective-order condition brings the produced list under a standing restriction on its use.
The tools for carrying out such orders are familiar ones. In Garner v. Wolfinbarger, a case about the availability of the attorney-client privilege, the court of appeals observed that a trial court can freely use in camera inspection and oral examination and freely avail itself of protective orders. It called the protective order a familiar device for preserving confidentiality in trade secret cases and others where the impact of revelation may be as great as in revealing a communication with counsel.14
An exam stem in which a company refuses to produce its formulas because they are “privileged trade secrets” invites two errors. The privilege claim fails, because no true privilege protects trade secrets. An answer that orders production with no conditions fails too, because it ignores what the rule is for. The expected analysis asks whether the issues can be fairly adjudicated without the information. If they can, the court may order that it not be revealed. If they cannot, disclosure comes with conditions on who sees the material, how it is used, and what happens to it afterward. The label “trade secret” describes the owner’s interest and does not carry the owner’s showing.
Step three is where most stems turn, because sensitivity is usually conceded and necessity is what a court actually weighs.
Sharing and Umbrella Orders
Commercial litigants commonly carry out the “specified way” of Rule 26(c)(1)(G) through a standing order that settles two questions.16 The first is who, beyond the parties and their lawyers, may receive what the producing party hands over. That question separates sharing orders from non-sharing orders. The second is how material comes within the order. That question produces the umbrella order. Both build on the conditional disclosure described above, and both give the producing party control over the audience at some cost to the requesting party.
Chudasama v. Mazda Motor Corp. frames the first question. Mazda moved for a protective order. The Chudasamas objected but said they would accept a “sharing” order that let them share the information with similarly situated plaintiffs, but not with anyone else.17 At a hearing, Mazda offered to stipulate to a sharing order if the plaintiffs would narrow their proposed definition of similarly situated plaintiffs, and the plaintiffs declined.18 The district court then adopted, nearly verbatim, the sharing order drafted by the plaintiffs’ counsel. The Eleventh Circuit noted that, whatever else could be said of that order, it still protected Mazda’s confidential information from disclosure to its competitors.19
The two proposals differ in one respect, the class of permitted recipients. An order that bars all onward sharing confines the material to the case. A sharing order widens the audience to a defined group of litigants with similar claims, and the fight in Chudasama was over how that group would be defined. A sharing order is therefore only as protective as its definition of the group. Fellow claimants may be adverse to the producer, but they are not its commercial rivals, and the drafting of the definition decides whether that line holds.
The second design question is how material comes within the order in a case that involves many documents. An umbrella protective order answers it by presumptively covering all material the disclosing party deems confidential, so the producer’s designation, not a ruling on each document, brings an item within the order.20 Parties frequently agree to such orders, and courts regularly issue them.21 The practical effect is that production can proceed under a standing promise of confidentiality without the court first ruling on each document.
Two mechanics give the order its force. To retain confidentiality if designated material is filed with the court or introduced into evidence, umbrella orders typically require that it be filed under seal.22 And a party who breaches a protective order is in contempt of court.23 Because the order is the court’s command rather than a private promise, a lawyer who passes designated documents to a reporter risks contempt. A private confidentiality agreement is a weaker tool. It binds only those who sign it, and it does not protect against a subpoena from someone who is not a party to the agreement. Litigants can instead jointly seek a protective order requiring the confidentiality of information exchanged in discovery and settlement negotiations.24
The two design questions are independent, so an exam answer should resolve them separately. An order can be an umbrella order that reaches whatever the producer designates and also a sharing order that lets the receiving party pass designated material to a defined group. The first feature fixes what is covered, and the second fixes who may receive it.
Chudasama also shows how seriously a court of appeals will take a producer’s reliance on such an order. The district court there struck Mazda’s answers and affirmative defenses, directed entry of a default judgment, ordered Mazda to pay the plaintiffs’ expenses and attorneys’ fees based in part on Rule 26(g)(3), and vacated the protective order it had entered earlier.25 Rule 26(g) is the initial control on discovery abuse. Like Rule 11, it certifies that papers are warranted by law and have a proper purpose, and it adds a warranty that disclosures are complete and correct when made.26 The Eleventh Circuit nonetheless held the sanctions order a clear abuse of discretion that could not stand under Rule 37(b)(2). It reasoned that the patent ambiguity in the discovery requests Mazda was compelled to answer, together with the district court’s failure to clarify Mazda’s obligations, was largely to blame for Mazda’s noncompliance, and that less onerous sanctions were available. Vacating the protective order, it said, was entirely unnecessary.27
The court added a reliance argument. The record showed that Mazda relied on the protective order when it provided a large share of its discovery responses, and the economic harm of competitor access to its proprietary information might exceed its liability in the suit.28 If the sanctions stood, Mazda and other large corporations might decide from the outset to disclose no proprietary information and accept a default. They would lose the lawsuit but keep the information out of the public record. A sanction meant to deter discovery abuse could then have the opposite effect.29 The reasoning runs in steps. The producer hands over material because the order promises a controlled audience. Vacating the order after production strips that protection from material already produced. Potential defendants would then learn that producing under a protective order is a risk, and they would withhold more.
None of this makes a protective order impregnable. A protective order is sometimes overcome by a subpoena in another case if the party seeking the discovery can demonstrate a substantial need or that the order was improvidently granted in the first action.30 Later sections take up those challenges. The narrower point here is that the Eleventh Circuit took a producer’s reliance on an entered order seriously.