On a summer afternoon in California, a wildfire breaks out in a remote canyon. Within minutes, and before the first 911 call is made, an artificial intelligence system will alert you to smoke, dispatch and direct firefighters to the scene. Meanwhile, when Los Angeles County residents seeking to rebuild their homes after the 2025 fire submit a permit application, an AI tool alerts them to errors and prompts for corrections before the permits reach the employees responsible for processing them. AI-enabled software cuts permit processing time in half.
These are unusual examples of public sector use of AI in California, but they are not hypothetical accounts. These examples demonstrate how public sector organizations can leverage AI to improve citizen protection and services.
The use of AI in public institutions has become a point of contention in the relationship between public institutions and the labor unions that represent civil servants.
Public sector unions are leveraging the increasingly negative public opinion toward AI to advance legislation to protect bargaining unit jobs and block, if not ban, the introduction and deployment of AI technologies and tools by public agencies.
Congress is currently considering several bills proposed by labor unions that seek to regulate and limit the use of AI by public authorities in the workplace, including the use of AI in employment decisions (AB 1898 and SB 947), the use of AI for workplace surveillance (AB 1883), and the use of AI to train future AI models (AB 2027).
Parliament Bill 2656 is one of a series of bills introduced by Labor aimed at regulating how public authorities can use AI and the limits that apply to its use. However, unlike other bills, AB 2656 takes a more labor-oriented approach, focusing on the relationship between public agencies and the unions that represent public employees.
New notification requirements
AB 2656 was framed by the bill’s sponsor, the Law Enforcement Union Association, as a transparency measure aimed at promoting accountability for the use of AI without slowing innovation, and would require public agencies to provide advance notice to unions about the use of AI.
Modeled after AB 339, which became law on January 1, 2026 and required public agencies to notify unions 45 days in advance of public agency contract solicitations that involve services within the “scope of work” performed by a representative employee, AB 2656 imposes similar notification requirements for public agencies’ use of AI.
Specifically, AB 2656 requires unions to be notified 45 days in advance if a public entity develops, purchases, implements, or uses AI to perform services that are within the “scope of work” of a representative employee. Unlike under AB 339, where it is more or less clear whether a contract decision pertains to services within an employee’s “scope of work,” such decisions are likely to be much more difficult in the context of AI use, where public agencies do not use technology or tools on behalf of employees, but instead enable employees to perform their jobs more efficiently and effectively.
redundant duties
There is little doubt that the increasing use of AI by public sector organizations is raising legitimate and important questions about the future impact of such technologies and tools on civil servants and their jobs. However, there is no doubt that the statutory collective bargaining system that applies to public agencies in California requires that when a public agency proposes a change to the terms and conditions of employment, it must notify the union and fully meet and consult with the union regarding its decision.
For example, under the Myers-Millius-Brown Act, California cities, counties, and special districts must notify unions about changes to their terms of employment and provide an opportunity to meet and discuss changes to the terms and conditions of employment, including those that affect an employee’s job duties or workload. Although there may be no precedent applying this well-established principle specifically to the use of AI by public bodies, the law is clear that where significant technological developments affect core terms and conditions of employment, public bodies must already notify trade unions and negotiate such decisions.
AB 2656 supplements, rather than replaces, these existing broader legal obligations, requiring public entities to notify unions before deciding to implement AI in a manner that necessarily changes fundamental aspects of an employee’s terms and conditions of employment.
Critics of AB 2656, including public sector organizations, cite the onerous nature of AB 339’s new notification requirements and argue that they would be made worse if Congress expanded them to include the use of AI. Critics also argue that such notification requirements are too broad and unnecessary, as existing legal obligations already require public authorities to notify employees when they decide to change their terms and conditions of employment.
Looking to the future
AB 2656 and other labor-backed AI bills may be amended, defeated, or vetoed by California Governor Gavin Newsom (D) in the future, but they provide a clear roadmap for labor’s new priorities regarding AI. A critical look at the bill reveals that things are more complex than buzzwords like “accountability” and “safeguards” suggest.
Like its predecessor, AB 339, AB 2656 may be a well-intentioned gesture toward transparency. But its real-world impact can be measured by lengthening schedules, consuming resources, and postponing decisions.
Lawmakers should consider whether additional layers of the process meaningfully protect workers or delay the implementation of tools that could benefit workers and the public they serve.
This article does not necessarily reflect the opinion of Bloomberg Law, Bloomberg Tax, Bloomberg Government, publisher Bloomberg Industry Group, Inc., or its owners.
Author information
Alexander Volberding is a partner at Liebert Cassidy Whitmore, where he advises public agencies in California on labor relations, collective bargaining, and workplace safety issues.
Gabriela Kamran is an associate at Liebert Cassidy Whitmore, where she advises public and educational institutions on labor relations, employment, and education law issues.
Interested in writing? Please review Author guidelinessend a pitch insights@bloombergindustry.com.
