abstract. There is a substantial literature on noncompete agreements and their adverse impact on employee mobility and innovation. But a far more common restraint in employment contracts has been underexplored: confidentiality agreements, sometimes called nondisclosure agreements (NDAs). A confidentiality agreement is not a blanket prohibition on competition. Rather, it is simply a promise not to use or disclose specific information. Confidentiality agreements encompass trade secrets, as defined by state and federal laws, but confidentiality agreements almost always go beyond trade secrecy, encompassing any information the employer imparted to the employee in confidence.

Despite widespread use, confidentiality agreements have received little attention. Many commentators view them as innocuous compared to noncompetes. However, confidentiality agreements that go beyond trade secrecy are not harmless. Leveraging an original dataset of confidentiality agreements in employment relationships disclosed in federal trade secret litigation, this Article argues that many of these agreements have the effect of noncompetes. They protect far more information than trade secret law does—including publicly available or generally known information, and information that trade secret law would classify as unprotectable “general knowledge, skill, and experience.” They prohibit use as well as disclosure of the covered information. Most provide for injunctions in the event of breach, and nearly half provide for payment of attorney’s fees and costs. And unlike most noncompetes, they almost never have geographic or temporal limitations.

The phenomenon of confidentiality agreements that “act like noncompetes” has not gone unnoticed. For example, the Federal Trade Commission (FTC) recently issued an unprecedented Notice of Proposed Rulemaking announcing its intention to effectively ban workplace noncompetes nationwide. The FTC also condemned what it calls “de facto non-compete clauses,” including overly broad confidentiality agreements. The FTC’s rulemaking has yet to move forward and is likely to be mired in legal challenges. Fortunately, this Article reveals that courts across the nation have already begun to invalidate confidentiality agreements that operate as de facto noncompetes. Regardless of whether the FTC ultimately succeeds in regulating these agreements, courts have the power and precedent to do so on their own.

Drawing on case law and prior proposals, this Article gives guidance going forward. It does not advocate for a blanket ban on confidentiality agreements. Rather, it contends that courts and other decision makers should treat confidentiality agreements that go beyond trade secrecy under a default rule of unenforceability, similar to how most jurisdictions treat noncompetes. The burden should be on the employer to prove that such agreements are reasonably related to protecting legitimately secret information and that they do not function like noncompetes.

authors. Camilla A. Hrdy is Professor of Intellectual Property Law, University of Akron School of Law; Affiliated Fellow, Information Society Project, Yale Law School. Christopher B. Seaman is Robert E.R. Huntley Professor of Law, Washington and Lee University School of Law. Thanks to W. David Ball, Victoria A. Cundiff, Nikola Datzov, Eric Goldman, Charles Tait Graves, Mark A. Lemley, David M. Levine, Michael Mattioli, Viva Moffat, Christopher Morten, Tim Murphy, Lisa Larrimore Ouellette, Donald J. Polden, Michael Risch, Elizabeth A. Rowe, Guy A. Rub, Sharon K. Sandeen, Evan Starr, and Deepa Varadarajan. Thanks also to participants at the 20th Works in Progress for Intellectual Property Scholars Colloquium (WIPIP 2023), and attendees of our talks for the Yale Information Society Project Law & Technology Speaker Series in April 2023 and the 2023 Conference on Empirical Legal Studies (CELS) at the University of Chicago Law School in October 2023. We are grateful for the invaluable research assistance provided by Nick Clarke, Lucie Fisher, John Gilmore, Anna Kennedy, and Sicai Zhu. Many thanks also to the editors at the Yale Law Journal, including Olu Oisaghie, Karina M. Shah, and Dena M. Shata.

Introduction

Confidential information can be protected under multiple legal regimes.1 In trade secret litigation—which typically involves a company suing a former employee for allegedly disclosing or using trade secret information—the most common state-law claims are for trade secret misappropriation and breach of contract.2 The reach of trade secret claims is limited to information that qualifies as a trade secret under state or federal law.3 The reach of contract claims, however, is not limited to the contours of trade secrecy. Rather, their scope is determined by the parties’ agreement. Contracts that go “beyond trade secrecy” can thus expand owners’ rights beyond those conferred by legislation.4

In the employment context, the most controversial such agreement is the noncompete. Noncompetes are contracts that prevent the recipient of information from competing following the exchange. If the defendant is an employee, the employee is typically prohibited from working at or starting a competing company in the same industry for a period of time.5 Noncompetes have been the subject of substantial public attention and controversy.6 They are per se unenforceable in some states, including California.7 In other states, noncompetes are treated as “restrictive covenants”8 and “restraints on trade”9 that must be “reasonable” in duration and geographic scope.10

The legal literature on noncompetes and their interaction with trade secret protection is vast.11 But noncompetes are not the only form of covenant appearing in civil trade secret disputes that can be used to expand the rights of employers over information. Another far more common restraint in employment relationships has largely fallen under the radar: confidentiality agreements, sometimes called nondisclosure agreements (NDAs).12 Unlike noncompetes, confidentiality agreements do not facially preclude an employee from competing or entering into an employment relationship. Instead, confidentiality agreements demand only that the recipient stay silent and not use, except as authorized, any information that the contract defines as “confidential.”13

With some exceptions,14 the general assumption has been that confidentiality agreements are innocuous compared to noncompetes. It is common to sign a confidentiality agreement before starting a new job.15 Even vociferous critics of noncompetes tend to characterize confidentiality agreements as “narrower restraints” that might be used to avoid the blunt prohibition of a noncompete.16 The common assumption is that, unlike a noncompete agreement, “a standalone NDA does not necessarily restrict an employee’s mobility options” because the employee “can still move to a competitor”; the employee simply cannot ever disclose or, “theoretically,” ever use, specific information obtained from their last employer.17 Courts across jurisdictions routinely give confidentiality agreements “more favorable treatment” than noncompetes.18 And confidentiality agreements are not typically subject to the same limitations that are applied to noncompetes.19 For instance, some states’ statutory regimes for noncompetes explicitly do not apply to confidentiality agreements.20 Overall, courts tend to apply a default rule of enforceability.21

Recently, some scholars and practitioners have begun to question the widespread assumption that confidentiality agreements in employment relationships are mostly harmless. They suggest confidentiality agreements might raise many of the same issues as noncompetes by effectively preventing employees from leaving and taking new jobs or starting their own companies.22 Moreover, because confidentiality agreements are not scrutinized by courts in the same way as noncompetes, they can theoretically last for the employee’s entire lifetime.23

This has real implications for workers. For example, a securities trader was recently bound by confidentiality provisions in his employment agreement that required him to “keep all Confidential Information in strictest confidence and trust” during and after his employment. The agreement defined “Confidential Information” to include—among other things—all “information, in whatever form, used or usable in,” “or relating to,” “analyzing, executing, trading and/or hedging in securities.”24 The agreement had no time limit. Its effect was to essentially prevent the employee “from trading in securities at all . . . for the remainder of his life.”25

Some policymakers are recognizing that broadly drafted confidentiality agreements can operate like noncompetes. For example, in 2021, the Uniform Law Commission (ULC) approved a Uniform Restrictive Employment Agreement Act (UREAA) that characterizes confidentiality agreements as “restrictive covenants” that are presumptively unenforceable unless they meet certain requirements.26 A few states, including Colorado, have passed legislation specifically addressing confidentiality agreements.27 Most recently, the Federal Trade Commission (FTC) has entered the scene. In an unprecedented Notice of Proposed Rulemaking issued in January 2023, the FTC notified stakeholders and the public of its intention to effectively ban noncompete agreements between employers and workers nationwide.28 The FTC announced that it would apply a “functional test” to determine what constitutes a noncompete. This test will encompass “non-disclosure agreements” that “are so unusually broad in scope that they function” as “de facto non-compete clauses.”29

So far, these concerns have occurred in an information vacuum because confidentiality agreements are not typically available for public access. Ironically, some confidentiality provisions specifically encompass the agreement itself.30 If academics, policymakers, and other interested parties do not know what these agreements look like, they cannot realistically assess their enforceability. As Orly Lobel aptly put it, these agreements are “enforceability TBD.”31

This Article is the first to provide a comprehensive picture of what confidentiality agreements look like when they are enforced in trade secret disputes. Leveraging a unique dataset of employment contracts publicly disclosed in federal trade secret cases brought in U.S. district courts, this Article identifies and analyzes confidentiality agreements that have been asserted in court alongside trade secrets.

To briefly summarize our results,32 we find that many confidentiality agreements generate such sweeping confidentiality obligations that it would be virtually impossible to work in the same field—let alone compete with a former employer—without breaching them. In other words, they have the functional effect of noncompetes. Several findings are particularly notable. First, the confidentiality agreements in our dataset protect far more information than trade secret law does. Most refer to “confidential” or “proprietary” information rather than limiting their scope to trade secrets. Second, many of the agreements encompass publicly available or generally known information, and nearly all of them implicitly extend to employees’ “general knowledge, skill, and experience.”33Third, these agreements are not just “nondisclosure” agreements because almost all of them prohibit use as well as disclosure of the covered information. Fourth, these contracts provide for injunctions in the event of breach—meaning that a worker can potentially be ordered, among other things, to stop using the covered information. Nearly half provide for payment of attorneys’ fees and costs. Finally, unlike noncompetes, they almost never have geographic or temporal limitations. They are usually written to last forever.34

If these provisions were unenforceable, the stakes would be lower. Employees who are sued for breaching these agreements would at least have the chance to win in court. But the historical wisdom among courts and commentators has been that confidentiality agreements are typically enforceable. Courts do not scrutinize them to the same degree as noncompetes.35 This is a serious problem. Although U.S. states have different approaches to noncompetes, all states subject noncompetes to at least some judicial scrutiny.36 If confidentiality agreements are given a free pass, then employers can simply get around legal restrictions on noncompetes by placing workers under perpetual, noncompete-like confidentiality obligations. Society would experience the negative effects discussed in noncompete law—restrictions on worker autonomy, competition, knowledge sharing, and cumulative innovation—but without any oversight at all.37

Fortunately, this Article reveals that this may be changing. There is a substantial body of recent case law from across the country in which courts have questioned the enforceability of confidentiality agreements that reach too far beyond trade secrecy.38 Most courts agree in dicta that confidentiality agreements can protect information that does not qualify as a trade secret.39 But when faced with cases in which plaintiffs try to protect non-trade-secret information—in particular, public or generally known information, or a worker’s general knowledge, skill, and experience—many courts do not enforce these agreements. Instead, they find them void for public policy or in direct contravention of the jurisdiction’s laws regulating noncompetes.40 Although California is the leader in refusing to enforce confidentiality agreements that act as de facto noncompetes,41 it is not alone. Even in jurisdictions that enforce reasonably tailored noncompetes, courts have begun striking down confidentiality agreements in employment relationships that go too far beyond trade secrecy, finding them to be violations of state statutes or common-law rules restricting enforcement of noncompetes.42

This case law is already having an impact, and it will be extremely important in a world in which policymakers do not or cannot address overreaching confidentiality agreements. For example, if the FTC fails in its attempt to police de facto noncompetes—which seems possible43—our research reveals that some courts are already doing so. Other courts can and should follow suit. Employers, in turn, will have to conform to these decisions, assuming they want their contracts to be enforceable. On the other hand, if the FTC succeeds, this case law will still be relevant because it will help regulators determine what it means for a confidentiality agreement to act like a noncompete. As we will show, one of the most important indicators is that the agreement goes too far beyond trade secrecy, hindering workers’ ability to use or share information that the law broadly intends to remain free.

This case law is evolving and imperfect. Not all courts scrutinize confidentiality agreements. The “enforceability default” is still the norm. Courts lack a uniform framework for evaluating their enforceability. To help courts resolve these issues, we provide a framework that is guided by principles articulated in recent case law and informed by our empirical results. In short, we argue that courts should treat confidentiality agreements in the workplace44under a default rule of unenforceability whenever they go beyond trade secrecy. Employers should have the burden to prove such agreements are reasonably related to the goal of preserving secrecy and do not have the effect of an unexpected noncompete agreement.45

Under this standard, many of the agreements reviewed in this Article would be unenforceable. But those that are tailored to protecting trade secrets or legitimately secret46 information could survive. A confidentiality agreement cannot, under our framework, cover public or generally known information. And it cannot create unexpected noncompete obligations—for instance, by preventing a worker from using their general knowledge, skill, and experience. We also discuss how courts should judge the duration of these agreements. Rather than a strict time limit, we advocate for a functional end date that looks to whether the information is still being kept secret. Employers should not be able to enforce a so-called “nondisclosure” obligation if the information has, in fact, already been disclosed.

The remainder of this Article proceeds as follows. Part I explains what confidentiality agreements are and how they can be used to protect information beyond trade secrecy. It also identifies the policy concerns implicated by confidentiality agreements that protect more than trade secret law allows.

Part II addresses the question of enforceability. It explains the commonly held view that confidentiality agreements are not subject to the same regulatory frameworks as noncompetes. But it then reveals case law from across jurisdictions in which courts have struck down confidentiality agreements that resemble noncompetes in effect. This Part also discusses recent legislation at the state level and recent proposals from the ULC and the FTC—all of which suggests that attitudes towards confidentiality agreements are rapidly changing.

Part III reveals the results of a unique empirical study that sheds light on what confidentiality agreements actually look like in practice. It shows that these agreements almost always protect more than trade secrets, and that many have the potential to act like noncompetes. Furthermore, it finds that very few of these agreements contain the durational or geographic limitations courts typically require for noncompetes. This Part also explains the study’s methodology and addresses its limitations.

Part IV provides a way forward for courts and policymakers. It explains the types of information that cannot be protected in workplace confidentiality agreements, as well as the types of information that can be. It assesses how courts should treat confidentiality agreements that go beyond trade secrecy. Ultimately, this Article does not advocate for a blanket ban on confidentiality agreements that go beyond trade secrecy. Rather, it provides a framework that courts can use to determine whether, and under what circumstances, confidentiality agreements that go beyond trade secrecy should be enforced. Our proposal aligns with the way some courts already assess these questions, and with the general direction taken by the FTC and the ULC, but it also clarifies the analysis and highlights the key issues that need to be considered.

Finally, the Conclusion advocates for balance. Companies should be able to use confidentiality agreements to protect real secrets, but they should not be able to place illegitimate and unnecessary restrictions on workers that were never bargained for. Otherwise, firms can simply circumvent long-established legal limits on noncompetes by using confidentiality agreements.