Noncompete agreements, which impose contractual limits on an employee’s ability to work after leaving a job, are regulated or banned in all states. But employers can potentially get around legal limitations on noncompetes by asking workers to sign confidentiality agreements that have similar functional effects. In a new article, Camilla A. Hrdy and Christopher B. Seaman provide empirical evidence that a significant number of employment agreements contain broad confidentiality provisions that place noncompete-like restrictions on workers.

A confidentiality agreement, in the workplace context, is a contract that prevents an employee from disclosing (and often from using) any information that the contract defines as “confidential.” Many workers sign a confidentiality agreement when starting a new job. But confidentiality provisions in employment agreements can be written so broadly that they effectively prevent an employee from working in the same field as their former employer after leaving a job. In other words, confidentiality agreements can act like noncompetes.  

Noncompetes are controversial. They explicitly limit competition and can have adverse impacts on employee mobility and innovation. Noncompetes are almost entirely banned in five states, including California, while a number of other states prohibit them for workers under a certain salary threshold or certain types of professions like primary care physicians. In the majority of states, a noncompete will only be enforced if it is justified by “legitimate business interests,” such as protecting trade secrets. Noncompetes must also be “reasonable” in time, geography, and scope of restricted activities. They don’t generally last more than a few years.

In contrast, confidentiality agreements have not traditionally been subject to these same legal limits. They have not been treated as restraints on trade. They are not banned and are not typically subject to reasonableness standards like noncompetes are. Confidentiality agreements that are limited to protecting trade secrets are generally not problematic. Trade secrets are specifically defined under federal and state law and protect only information that is not generally known or readily ascertainable, and that the employer took reasonable measures to keep secret. Furthermore, trade secrecy does not prohibit workers from using or sharing public information, general knowledge, or an employee’s general knowledge, skill, and experience.

The problem is that many confidentiality provisions are not limited to protecting just trade secrets. Rather, they define “confidential” information much more broadly. The word “confidential,” on its own, just indicates that the information is deemed confidential by the parties, and that the recipient is under a duty of confidentiality. The word is not necessarily synonymous with trade secrets.

In practice, confidentiality provisions can be so sweeping in scope that they make it difficult for employees to avoid breaching them after they leave a job. For example, one staffing agency required an employee to sign a confidentiality agreement that provides that, after leaving the job, the employee cannot “use or divulge, disclose or communicate…any confidential information of any kind, nature, or description concerning any matters affecting or relating to the business of the Company,” such as customer lists, sales and manufacturing information, financial information, code books, or “any other confidential information of, about, or concerning the Business of the Company, its manner of operation, or other confidential data of any kind nature or description.” The agreement contains no exclusions, even for information that is generally known. The agreement has no geographic or temporal limitations.  

If confidentiality agreements can be written so broadly that they have the functional effect of noncompetes, then companies can get around the restrictions on noncompetes by imposing perpetual, noncompete-like confidentiality obligations on workers. Our study of confidentiality agreements, which we discuss below, suggests that many confidentiality provisions are indeed being drafted to be very broad and may have some of the same functional effects as noncompetes, while also being written to last forever and apply world worldwide.

This development should be concerning to pretty much everyone. Even those who support enforcing reasonable noncompetes should oppose overbroad confidentiality agreements of the type described above. Society would get all the policy harms that are recognized in noncompete law—restrictions on competition, less opportunities for knowledge sharing and cumulative innovation, and burdens on worker autonomy—but without any meaningful regulation or oversight. Most importantly, these agreements are functionally perpetual, which is not justifiable even under the most “pro-noncompete” frameworks.

We are not alone in questioning employers’ regular use of confidentiality provisions. Numerous scholars, such as Rachel Arnow-Richman, Orly Lobel, Sharon Sandeen, Evan Starr, and Deepa Varadarajan, among others, have expressed concern that employees might be subject to overbroad contractual provisions that go beyond the protections of trade secret law and that reach into noncompete territory. But to date, the general consensus on confidentiality agreements has remained positive. They are often cited as harmless alternatives to noncompetes that are not different in kind from trade secrets.

Fortunately, this view is changing in some quarters. In January 2023, the Federal Trade Commission proposed a nationwide ban on noncompetes agreements in employment contracts, including “de facto noncompetes.” The agency’s primary example of a de facto noncompete is “a non-disclosure agreement between an employer and a worker written so broadly it effectively precludes the worker from working in the same field after the conclusion of the worker’s employment with the employer.”

With a potential ban on the horizon, our article makes two contributions. First, we provide empirical evidence that some confidentiality agreements are indeed written to be very broad and have at least the potential to act like noncompetes. Using an original dataset of noncompete agreements disclosed in federal trade secret litigation between May 2016 and May 2017, we identified and coded 446 confidentiality agreements entered between employers and workers for a variety of criteria. The main take-away from our data is that confidentiality agreements usually go beyond trade secrecy, and in many cases have characteristics that make them act like noncompetes.

First, while many of the agreements use the term “trade secrets,” the vast majority use broader terms like “confidential” or in some cases “proprietary.” And in tension with trade secret law, about 40% of these agreements do not expressly exclude information that is public or generally known in the industry. Second, the vast majority (88%) prohibit use as well as disclosure of the protected information. This is significant because it makes the agreements more like noncompetes in effect. Employees can be liable for breach of contract even if they don’t deliberately share the information. Third, the confidentiality agreements in our dataset usually provide that the employer can seek injunctive relief, meaning they can stop an employee from using covered information altogether. Again, this makes these agreements more like noncompetes in effect. An employee can literally be stopped from competing. Finally, over 90% of confidentiality agreements in our data set contained no durational limit, meaning that they effectively last for the employee’s lifetime. Furthermore, almost none of these confidentiality provisions had any geographic limit, making them effectively worldwide in scope.

The article’s second contribution is to dive into the case law regarding enforcement of employer/employee confidentiality agreements. We show that state and federal courts in jurisdictions as diverse as Arkansas, California, Illinois, South Carolina, and West Virginia are pushing back against overbroad confidentiality provisions by striking down some of the most broadly-drafted provisions, often under their state’s noncompete laws. While a welcome development, this bottom-up approach is piecemeal at best, and reversible at worst.

With all this in mind, how should courts and regulators address confidentiality and nondisclosure provisions in future and decide when they are enforceable, and when they are not? We contend that when a confidentiality agreement in an employment contract protects more than trade secrets, it should be treated similar to a noncompete and be presumptively unenforceable. The employer can rebut this presumption, but it should have the burden to show two things: First, the employer must demonstrate that the information at issue is currently a trade secret or at least is not public or generally known at the time of enforcement. Second, the employer that the agreement is not so excessively broad that it has the effect of a noncompete, leaving the worker free to utilize their talents and training to accept employment.

Notably, our approach remains valid regardless of the status of the FTC’s proposed rule that would effectively ban noncompetes nationally. If the FTC ban is enacted as proposed, our analysis will help elucidate what qualifies as a de facto noncompete. It seems clear from courts’ decisions that the problem is not confidentiality agreements per se, but confidentiality agreements that prevent workers from using general knowledge, skill, and experience that they need to do their jobs. The FTC should be clear that confidentiality agreements are far more likely to be treated as de facto noncompetes when they go beyond trade secrecy. On the other hand, if the FTC’s proposal is modified or struck down in court, our article shows that the case law should continue moving in the direction of regulating confidentiality agreements that act like de facto noncompetes by placing the burden on employers to prove these are legitimate.

Articles represent the opinions of their writers, not necessarily those of ProMarket, the University of Chicago, the Booth School of Business, or its faculty.