California Bills Would Require Human Review of AI Firings and 90-Day Notice for AI Layoffs

California lawmakers introduced two bills yesterday that would significantly restrict how employers use artificial intelligence to make employment decisions. The coordinated effort by the California Labor Federation targets AI-driven job losses with a two-pronged approach: requiring human oversight when AI is used to fire or discipline workers, and mandating extended advance notice before conducting mass layoffs driven by automation. If enacted, the measures would impose some of the strictest AI employment regulations in the nation. Here’s what California employers (and multistate employers watching the trend) need to know about these February 2 proposals.

SB 947: The “No Robo Bosses” Act Returns

State Senator Jerry McNerney’s SB 947 is a revised version of legislation Governor Gavin Newsom vetoed last year. We covered the original proposal and the governor’s veto in detail in 2025.

The core provision remains: employers would be prohibited from using “automated decision systems” (ADS) as the sole basis for making consequential employment decisions about current employees, specifically decisions to discharge or discipline workers. The bill defines ADS broadly as any “computational process, including one derived from machine learning, statistics, or other data processing or artificial intelligence techniques, that makes a decision or facilitates human decision making.”

What’s Changed From Last Year

The 2026 version addresses some of Governor Newsom’s concerns while adding new restrictions:

• Removed: The requirement to provide advance notice to workers before deploying AI systems for employment decisions. Newsom cited this as overly burdensome in his veto message.
• Changed: Now requires only retroactive notice after an AI system is used to make or substantially assist in a discharge or discipline decision.
• Added: A new prohibition on using ADS systems that attempt to predict employees’ future behavior. This ban targets emerging “predictive analytics” tools that claim to forecast worker performance, turnover risk, or misconduct likelihood.
• Retained: The fundamental requirement that a “natural person” must make the final decision when AI is involved in firing or disciplining workers. Employers cannot delegate these decisions entirely to automated systems.
• Retained: Default contractual language establishing that ADS cannot be the sole basis for discharge or discipline decisions, though this language can be modified through collective bargaining.

Violations would be enforceable by the California Labor Commissioner, who could impose civil penalties and order reinstatement or other remedies. The bill also provides that a worker or their exclusive representative can bring a civil action for damages, including punitive damages.

SB 951: 90-Day Notice for AI-Driven Mass Layoffs

State Senator Eloise Gómez Reyes introduced SB 951 as a companion measure focused on mass layoff events caused by automation or AI implementation. The bill would amend California’s existing Worker Adjustment and Retraining Notification (Cal-WARN) Act to create enhanced notice requirements specifically for AI-driven layoffs:

• Extended Notice Period: Employers would need to provide 90 days’ advance notice before conducting mass layoffs “due to the use of artificial intelligence or automation” – 30 days longer than the standard 60-day Cal-WARN requirement.
• Lower Threshold: The AI-specific notice requirement would trigger at 25 affected workers or 25% of the workforce (whichever is less), compared to standard WARN’s threshold of 50 or more employees laid off within a 30-day period.
• Job Bidding Rights: For employers with 100+ employees, affected employees would have the right to bid on any available positions within the company during the notice period, giving them an opportunity to avoid layoff through internal transfer.
• State Reporting: Employers would need to notify the California Employment Development Department about AI-driven layoffs, creating a state database to track automation-related job losses.
• Expanded Coverage: Unlike federal WARN, which only covers employers with 100+ employees, SB 951 would apply to employers with 75 or more employees, consistent with existing Cal-WARN thresholds.

What This Means for California Employers

If either or both bills become law, California employers would face significant new compliance obligations when using AI in employment decisions. Here are the key practical implications:

Immediate AI Audit Recommended

Employers should inventory all systems currently in use that could qualify as “automated decision systems” under these bills. The definition is broad enough to encompass:

• Applicant tracking systems that rank or score candidates
• Performance management platforms that flag employees for discipline
• Workforce analytics tools that identify layoff candidates
• Predictive models that assess flight risk or performance trajectories
• Scheduling algorithms that could affect hours or shifts

Don’t limit your review to tools explicitly marketed as “AI.” The statutory definition covers any computational process that uses data processing techniques to make or facilitate decisions.

Human-in-the-Loop Documentation

Under SB 947, employers would need to demonstrate that a natural person made the final decision in any discharge or discipline case where AI played a role. This means:

• Document human review: Create records showing a person reviewed AI recommendations and applied independent judgment
• Train decision-makers: Ensure managers understand they cannot simply rubber-stamp AI outputs
• Establish override procedures: Build processes for humans to disagree with or override AI recommendations
• Preserve evidence: Maintain records that could prove human involvement if challenged by the Labor Commissioner

Simply having a human click “approve” on an AI-generated termination recommendation likely won’t satisfy the bill’s requirement for meaningful human decision-making.

Retroactive Notice Obligations

While the revised bill eliminates advance notice requirements, employers would still need to provide notice after using AI for discipline or discharge decisions. The bill’s language on retroactive notice isn’t fully detailed yet, but employers should prepare to:

• Inform employees when AI played a role in a termination or discipline decision
• Explain what the AI system analyzed or recommended
• Describe how the human decision-maker used or departed from the AI’s output

This creates potential complications in termination meetings and unemployment hearings, where employers will need to disclose their use of AI tools.

Extended Layoff Planning

SB 951’s 90-day notice requirement would force employers to plan AI-driven workforce reductions much further in advance. Key considerations:

• Earlier strategic decisions: Technology implementation timelines would need to account for the extended notice period
• Competitive disadvantage: The requirement to announce layoffs 90 days in advance could harm employers in competitive markets, potentially explaining why Newsom has been reluctant to impose excessive AI regulations
• State reporting burden: Employers would need systems to track and report AI-driven layoffs to state agencies
• Internal bidding processes: HR would need to manage the new job bidding rights for affected workers

The Political Landscape: Will These Bills Become Law?

The path to enactment is uncertain. Several factors will influence whether these bills reach the governor’s desk and, if so, whether Newsom signs them. On the one hand, these bills appear to be a top priority for Labor, as the California Labor Federation is putting significant political capital behind these bills after losing ground in 2025. And SB 947’s removal of advance notice requirements addresses Newsom’s specific objection from last year.

On the other hand, these bills will face tech industry opposition, as they will likely argue the bills could make California less competitive and create compliance burdens. Moreover, the governor has consistently resisted “overregulating” AI, viewing California’s tech industry as crucial to the state’s economy (and especially as he eyes a bid at the White House in 2028).

Finally, you should pay attention to the White House drive to stop states from regulating AI. We expect the Department of Justice to file lawsuits against California and other states in a bid to stop them from enforcing these types of laws, and President Trump will continue to push Congress in an attempt to pass a national law preempting states like California from regulating AI.

Conclusion

We will continue to monitor the progress of these bills and other AI employment legislation in California and nationwide. Make sure you are subscribed to Fisher Phillips’ Insight System to get the most up-to-date information. If you have questions about AI governance, California employment law compliance, or how these proposals might affect your organization, contact your Fisher Phillips attorney, the authors of this Insight, or any member of our AI, Data, and Analytics Practice Group or any of our California offices.