Employers now deploy AI-powered productivity monitors, keystroke loggers, algorithmic pace-setters, and real-time sentiment analysis tools at a scale that would have required an army of supervisors a generation ago. Whether any of this is lawful under the National Labor Relations Act depends on a doctrinal intersection that neither the Board nor the courts have fully mapped.
I. Section 7 and the Contours of Protected Concerted Activity
29 U.S.C. § 157 — Section 7 of the NLRA — guarantees employees "the right to self-organization, to form, join, or assist labor organizations, to bargain collectively through representatives of their own choosing, and to engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection," as well as the right to refrain from such activities. Section 8(a)(1) makes it an unfair labor practice for an employer to "interfere with, restrain, or coerce employees in the exercise of the rights guaranteed in Section 7."
Activity is "concerted" when it is engaged in with or on the authority of other employees, not solely by and on behalf of an individual. It is "protected" when it concerns employees' interests as employees. The practical reach of § 7 is thus broad: conversations about wages, discussions about working conditions, complaints about supervisors, and organizing activity all fall within it. Crucially for surveillance purposes, the preparation for concerted activity — gathering information, drafting communications, identifying co-workers who might join a grievance — is also protected. An employer who monitors its employees in ways that chill this preparatory activity violates § 8(a)(1) even if no formal organizing campaign is underway.
II. NLRB General Counsel Memorandum GC 23-02
On October 31, 2022, NLRB General Counsel Jennifer Abruzzo issued General Counsel Memorandum GC 23-02, "Electronic Monitoring and Automated Management of Employees Interfering with the Exercise of Section 7 Rights." The memo is a significant framing document even though it represents the GC's proposed approach, not Board precedent.
GC 23-02 catalogs the surveillance technologies at issue: wearable devices with biometric sensors, GPS tracking, radio-frequency identification badges, cameras, keyloggers, screenshot-capturing software, audio monitoring applications, and — critically — AI-driven algorithmic management tools that set work pace, evaluate productivity, and issue discipline based on data aggregation. The GC's core observation: "Close, constant surveillance and management through electronic means threaten employees' basic ability to exercise their rights" by creating a chilling effect on the communications and associations that concerted activity requires.
The GC proposed a new framework for evaluating electronic monitoring under § 8(a)(1): an employer presumptively violates the Act where its surveillance and management practices, viewed as a whole, would tend to interfere with or prevent a reasonable employee from engaging in Section 7 activity. If the employer demonstrates that its practices are narrowly tailored to a legitimate business need — meaning the need cannot be met through less rights-invasive means — the Board would balance respective interests. Even where employer interests prevail, the GC proposed mandatory disclosure: employers would be required to inform employees of the monitoring technologies deployed, the reasons for their use, and how the collected information is used.
Critical practice note: GC 23-02 was formally rescinded in February 2025 by NLRB General Counsel William Cowen, along with numerous other Abruzzo-era memos. The rescission does not return the law to the pre-2022 status quo — the framework proposed in GC 23-02 was based on existing Board law and those precedents remain. However, the likelihood of the Board adopting GC 23-02's proposed new framework during the current administration is substantially reduced. Practitioners advising clients should focus on the existing precedent infrastructure, not the rescinded memo's proposed standards, while remaining aware that a future administration could revive the framework.
III. The Stericycle Work Rules Standard
The Board's August 2023 decision in Stericycle, Inc., 372 NLRB No. 113 (2023) significantly tightened the standard for evaluating facially neutral employer work rules under § 8(a)(1). Stericycle replaced the employer-friendly categorical approach of Boeing Co., 365 NLRB No. 154 (2017), with a new two-step analysis.
Under Stericycle: (1) a rule is presumptively unlawful if, when reasonably interpreted by an economically dependent employee contemplating Section 7 activity, the rule "has a reasonable tendency" to chill protected conduct — the employer's intent is irrelevant; (2) the employer may rebut the presumption by demonstrating both that the rule advances a "legitimate and substantial business interest" and that it cannot be replaced with a more narrowly tailored alternative.
Stericycle arose in the context of employee handbook provisions governing confidentiality and civility, but its logic extends directly to AI monitoring policies. An AI productivity surveillance policy that informs employees their communications are monitored for content — even if facially neutral in its stated purpose — has a reasonable tendency to chill employees from discussing wages, working conditions, or organizing activity in workplace communication channels. Under Stericycle, that chilling tendency triggers the presumption of unlawfulness, and the employer must then demonstrate both a legitimate business need and the absence of less restrictive alternatives.
Importantly, Stericycle's fate is also uncertain under the current Board, which has not yet addressed whether to apply, modify, or revisit the standard. The decision remains current Board precedent but should be flagged as potentially subject to modification.
IV. Application to AI Productivity Monitoring
AI worker surveillance tools generate an array of specific § 7 concerns that practitioners must analyze with precision.
Chilling effect from comprehensive monitoring. Courts and the Board have long held that an employer violates § 8(a)(1) by photographing employees engaged in picket activity or reviewing social media to identify union supporters. AI monitoring systems that parse employee communications for sentiment related to management criticism, compensation complaints, or organizing-related keywords operate on a larger scale and with greater precision. Under existing § 8(a)(1) doctrine — independent of GC 23-02's proposed framework — an employer that deploys monitoring technology for the purpose of detecting or suppressing protected activity commits an unfair labor practice. Discovery in Board proceedings should focus on whether AI monitoring vendors were asked to flag union-related communications and whether monitoring was initiated or intensified in response to organizing activity.
Algorithmic management and discipline. Warehouse and gig-economy employers increasingly use AI systems to set productivity quotas, penalize workers who miss targets, and generate automated discipline recommendations. If quota systems, break-monitoring tools, or other automated management practices disproportionately penalize workers for taking time to speak with co-workers about working conditions — as opposed to taking personal breaks — they may have a reasonable tendency to chill protected activity under Stericycle. The analysis requires showing that the algorithmic system treats concerted-activity-related behavior less favorably than comparably non-protected behavior.
Surveillance of off-duty activity. The Board has recognized that employers generally may not monitor employees' off-duty activities, including union meetings and communications. AI-based monitoring extending to employees' personal devices, off-duty social media accounts, or location data during non-work hours raises distinct § 7 concerns and may also implicate state electronic-surveillance and wiretapping laws independent of NLRA analysis.
Installation of monitoring in response to protected activity. Installing new monitoring technology in response to union organizing, or disclosing existing monitoring in a way designed to create the impression of surveillance, has been found unlawful under existing Board precedent without any need for GC 23-02's proposed framework. This is one of the most reliably unlawful forms of AI surveillance: if an employer rolled out AI keystroke monitoring one week after employees signed union authorization cards, the timing is powerful circumstantial evidence of retaliatory motivation.
V. Practice Notes for Union-Side and Employee Counsel
Filing ULP charges. Unfair labor practice charges must be filed with the NLRB regional office within six months of the alleged violation. 29 U.S.C. § 160(b). Practitioners representing employees or unions should identify the specific conduct alleged to chill § 7 activity, the monitoring technology deployed, and any causal connection to subsequent adverse employment action. Board proceedings move faster than federal court — pre-election injunctions under § 10(j) remain a powerful tool when surveillance is part of a broader anti-organizing campaign.
Section 8(a)(3) retaliation. Where an employee is disciplined or discharged following AI-surveillance-generated performance findings, and the timing or targeting suggests the protected activity was a motivating factor, § 8(a)(3) — which prohibits discrimination based on union membership or protected activity — provides an independent theory. The Wright Line test (251 NLRB 1083 (1980)) requires the General Counsel to show that protected activity was a motivating factor; the burden then shifts to the employer to show it would have taken the same action absent the protected activity.
Employer-side counseling. Employers deploying AI surveillance should: (a) document the business rationale for the monitoring system before deployment; (b) provide advance notice to employees of the monitoring technology and its purposes — disclosure both satisfies the good-faith information-sharing principle and is a defense to a chilling-effect claim; (c) limit access to the collected data on a need-to-know basis; and (d) conduct a pre-deployment review of whether the system is designed in a way that could disproportionately flag protected activity.
VI. Open Questions
Board adoption of GC 23-02's framework. Under a future NLRB General Counsel sympathetic to the GC 23-02 approach, the Board could adopt a comprehensive algorithmic management framework without any need for new legislation. The legal infrastructure supports it; whether a future Board will exercise the authority is a political question.
Mandatory bargaining over surveillance technology. In unionized workplaces, employers may have a mandatory bargaining obligation before implementing AI monitoring systems that affect "wages, hours, and other terms and conditions of employment" under § 8(d). The Board has treated employer monitoring policies as mandatory bargaining subjects in specific contexts. Unions negotiating CBA successors should insist on notice-and-consent provisions governing new surveillance technology.
State law. Several states — notably California, Illinois, and Washington — impose independent notice and consent obligations for workplace electronic monitoring that operate regardless of NLRA doctrine. An AI surveillance program that passes NLRA scrutiny may nonetheless violate state law, and state-law remedies may include statutory damages not available in Board proceedings.
VII. Closing
Section 7's protection of concerted activity is not a historical artifact — it applies with full force to AI monitoring tools that create documented chilling effects on workplace communication. GC 23-02, though rescinded, accurately identified the doctrinal infrastructure supporting greater scrutiny. Stericycle's presumptive-unlawfulness standard for work rules tightens the employer's burden to justify monitoring policies that touch employee communications. Practitioners on both sides should approach AI worker surveillance as the active NLRA battleground it has become, track Board precedent closely, and not assume that the absence of a comprehensive Board decision means the law is settled.
Talk to Yates Anderson
If you are litigating a matter in this area — or weighing whether to — the working analysis above only goes so far. Request a case evaluation and a Yates Anderson attorney will respond within one business day.
Informational only. Not legal advice. No attorney-client relationship is created by reading this post. Consult a licensed attorney in your jurisdiction.