National Labor Relations Board Sets Limits on Severance Agreement Provisions

Last month, the National Labor Relations Board (NLRB) made a decision that changed the game regarding severance agreements. The ruling confirmed that overly broad confidentiality and non-disparagement clauses have a “clear chilling tendency” and that offering agreements with such provisions violates the law. However, the NLRB did not ban severance agreements outright, as long as they are narrowly tailored.

Severance agreements are still allowed, but with certain limitations as confirmed by the NLRB decision. Employers must take care to draft severance agreements that do not include overly broad confidentiality and non-disparagement provisions. These types of clauses create an environment in which employees may feel that they cannot speak out against their employer, even if they have legitimate concerns about the conditions of their employment or the workplace culture.

Overly broad confidentiality and non-disparagement provisions have a “clear chilling tendency.” The NLRB has found that confidentiality and non-disparagement provisions in severance agreements are unlawful when they prevent employees from exercising their rights under the National Labor Relations Act (NLRA). These provisions are overly restrictive and prevent employees from speaking out about legitimate concerns, which can be protected under the NLRA.

Employers should take caution when drafting severance agreements. According to the NLRB decision, offering agreements with overly broad confidentiality and non-disparagement provisions violates the law. This is because such provisions create a legally binding agreement that reinforces the employer’s efforts to restrict an employee’s protected rights.

The good news is that confidentiality and non-disparagement provisions in severance agreements are still permitted – as long as they are narrowly tailored. The NLRB decision reiterated the importance of narrowly tailoring such provisions, which should include a temporal limitation and should be conveyed as clearly as possible to avoid confusion.

According to the NLRB’s latest memo, employers should avoid using language that may interfere with their employees’ exercise of NLRA rights in any form of communication. This includes handbooks, policies, forms, and other less formal agreements such as severance agreements. It is suggested that supervisors should not discourage employees from discussing the terms of their severance agreements, nor should they retaliate against those who attempt to do so.

Employers need to be aware that the NLRB’s decision has a retroactive effect, i.e., it applies to both existing and new severance agreements. If they are found to have violated these provisions, employers may be subject to legal penalties or may be required to rewrite the provisions in question. Therefore, it is advisable for companies to review their existing severance agreements to ensure compliance with the new guidelines.

The importance of tailoring agreement language cannot be overstated. It is crucial for employers to carefully draft and tailor the language used in severance agreements. The NLRB decision serves as a reminder that boilerplate savings clauses or disclaimers will not necessarily cure overly broad provisions. Employers must reassure employees that they still have the ability to exercise their NLRA rights without fear of retaliation, censorship, or any other undue limitation.

The recent decision of the NLRB on severance agreements has established clear limits on confidentiality and non-disparagement provisions that are too broad. To avoid legal repercussions and difficulties, employers should meticulously draft the language used in severance agreements to restrict their employees’ NLRA rights as narrowly as possible. Employers should also avoid using language that could interfere with the exercise of NLRA rights in any other communications with their employees.

Explore more

Ethereum Tests Glamsterdam Upgrade Amid Market Volatility

The activation of the Glamsterdam upgrade on the Sepolia testnet marks a critical phase in Ethereum’s infrastructure scaling as the network tests a gas limit increase from 60 million to 200 million. This substantial expansion of the gas limit represents a calculated gamble on the robustness of current hardware, aimed at accommodating a new wave of high-throughput decentralized applications. While

How to Design and Optimize AI Prompts for Production

The shift from experimental chatbots to high-scale enterprise intelligence systems in 2026 has transformed prompt engineering from a creative writing exercise into a disciplined branch of software engineering. The most effective production prompts use structural separation to distinguish between trusted system instructions and untrusted content from user inputs or retrieved documents. When an application processes thousands of model calls against

What Are the Best Email Marketing Tools for SMBs in 2026?

Small businesses often choose Constant Contact because it offers an extensive library of templates and specialized tools for managing event registrations and ticketing directly through emails. However, the broader landscape of digital outreach has shifted significantly, transforming email from a simple messaging tool into a sophisticated infrastructure for revenue growth and long-term customer retention. In 2026, the success of a

EY Breach Exposes Goldman Sachs and Man Group Client Data

Administrative IT tickets used for routine tax services inadvertently served as a repository for sensitive client data that was eventually stolen by hackers. This security failure at Ernst & Young (EY) has sent ripples through the financial sector, as it compromised the personal information of high-net-worth individuals associated with Goldman Sachs and the London-based hedge fund Man Group. While these

New Phishing Campaign Impersonates AI Tools to Steal MFA Codes

The campaign exploits the established trust that advertising agencies place in AI tools to bypass multi-factor authentication protocols that were previously considered secure. This sophisticated operation, identified in late 2026, represents a significant shift in the threat landscape, moving away from generic banking lures and toward the highly specialized tools used by modern marketing professionals. By impersonating platforms such as