Yes. California public workers can challenge AI-driven job cuts under state law, mainly through human-oversight rules and union bargaining rights. An automated decision system means software that helps decide discipline or firing, and the state now limits firing based only on that output.
According to Crowell & Moring via Legal 500, SB 947 bars employers from relying solely on those systems to fire or discipline workers Crowell & Moring analysis via Legal 500. The law is called the No Robo Bosses Act and starts July 1, 2027. It covers public and private workplaces.
Table of Contents
- What SB 947 stops
- What rights do workers get in an AI firing?
- How can unions challenge AI job cuts?
- What help and limits should displaced workers expect?
What SB 947 stops
An agency cannot let software alone decide that a worker should be punished or terminated. A person must meaningfully stand behind the action. That rule targets robo-firings where a score, flag, or recommendation becomes the decision. The protection applies to discipline and termination, not every use of AI at work. Scheduling help, drafting tools, or research aids are different.
The key question is whether the AI output primarily drove the job loss. A separate law, Assembly Bill 1883, limits emotion and neural-data surveillance. Atkinson, Andelson, Loya, Ruud & Romo reports it takes effect Jan. 1, 2027. It restricts bossware-style monitoring by California public employers.
What rights do workers get in an AI firing?
When an employer primarily relies on such a system, it must add human corroboration. According to Atkinson, Andelson, Loya, Ruud & Romo, the employer must also give written notice, provide access to relevant data, and faces Labor Commissioner enforcement with $500 penalties per violation AALRR alert on California AI workplace laws. Those duties create a paper trail workers can review. That trail matters for a challenge.
Ask what system was used and what data it relied on. Keep the notice, ratings, productivity scores, attendance flags, and names of human reviewers. A worker can then test the decision for error or bias. Wrong inputs, missing context, and unreviewed automation are practical attack points. A union representative or employment adviser can help frame that dispute.
📨 Get Free California Guides Alerts
Free · No spam · Unsubscribe anytime
How can unions challenge AI job cuts?
Many fights will run through labor bargaining, not individual lawsuits. Agencies must still consider whether AI adoption triggers meet-and-confer duties over implementation and effects. According to Bloomberg Law, challenges are brought as unfair-practice charges to the Public Employment Relations Board Bloomberg Law on union notice duties. A proposed 45-day union-notice rule for generative AI doing represented work did not become law.
AI Laws by State notes Assembly Bill 2656 was vetoed Sept. 30, 2026 as better handled through bargaining. So there is no automatic statewide notice period. Workers facing agency AI changes can take these steps:.
- Ask the union whether the agency gave notice and bargaining opportunity
- Request the AI policy, impact statement, and effects on duties or staffing
- File a grievance promptly and preserve deadlines
- Support a bargaining demand for retraining, reassignment, or transfer rights
What help and limits should displaced workers expect?
Executive Order N-6-26 directs state agencies to build support for AI-displaced workers. The Office of Gov. Gavin Newsom says that support includes severance standards, unemployment-insurance help, transition support, retraining, and hiring tracking. Small-business effects are also part of the order. Bargaining has already produced stronger local protection in one city.
Silicon Report describes IFPTE Local 21 members voting 94.1% on Oct. 22 to authorize a strike over municipal algorithms, then winning a tentative three-year deal favoring retraining or reassignment over termination. That outcome shows what contract language can add beyond minimum state rules. Statewide job-loss data did not show a broad surge as of last spring. The UCLA California Center for Population Research describes the AI-Unemployment Tracker, built with the Governor's Office, Employment Development Department, and UC California Policy Lab, as finding no statewide surge in claims from highly AI-exposed occupations as of May 2026. Workers should therefore document their own case rather than rely on broad trends.