California signs AI worker protection laws requiring human review, layoff disclosure and surveillance limits

On September 30, 2026 Governor Gavin Newsom signed a package of AI‑focused statutes that target both workers and consumers. The legislation marks the first U.S. state effort to embed human‑review requirements into employment decisions, to make AI‑induced workforce reductions transparent, and to limit the use of AI‑driven surveillance in sensitive areas such as restrooms. The measures complement a broader California AI framework that also touches health care, higher‑education training, public procurement, legal practice, deep‑fake impersonation and gene‑synthesis screening.

Key provisions for employment decisions

• SB 947 – bars employers from relying solely on AI for disciplinary or termination actions.
• SB 951 – creates a duty for human review of any AI system that could affect an employee’s job status.
• AB 1331 – requires disclosure when a mass layoff, relocation or termination is caused by an AI system.
• AB 1883 – adds documentation obligations for AI‑driven workforce actions.

Together, SB 947 and SB 951 prevent a purely algorithmic approach to firing or disciplining staff. Under SB 947, an employer may use AI as a tool, but a qualified human must make the final decision. SB 951 expands the requirement, mandating that any AI system with the potential to affect employment outcomes be subject to a documented human‑review process before it can be acted upon. The statutes do not ban AI in the workplace; they simply require that a person, not a machine, be the ultimate arbiter of adverse employment actions.

Mandatory disclosure for AI‑driven workforce changes

AB 1331 and AB 1883 address the growing anxiety that AI could silently trigger large‑scale job cuts. The law obliges employers to disclose, in writing, any mass layoff, relocation or termination that is directly caused by an AI system. The disclosure must be provided to affected employees and to the state labor agency, creating a transparent record of AI‑induced workforce disruptions. The legislation does not define a specific threshold for what constitutes a “mass” event, leaving the determination to the agencies that will enforce the rule.

Limits on workplace surveillance

The new statutes also place explicit limits on AI‑enabled surveillance. Employers may no longer deploy AI monitoring tools in private areas such as bathrooms, locker rooms or other spaces where a reasonable expectation of privacy exists. The law further restricts continuous AI‑based tracking of employee movements or biometric data unless a clear, job‑related purpose is documented and the employee is given an opt‑out option where feasible. These provisions aim to balance legitimate safety or productivity concerns with fundamental privacy rights.

Broader AI regulatory context in California

The worker‑focused bills sit alongside other California initiatives announced on the same day. In health care, the state requires AI tools that influence diagnosis or treatment to undergo independent validation and to be disclosed to patients. Higher‑education institutions receiving state funds must embed AI ethics and safety modules into curricula. Public procurement rules now mandate that AI systems purchased by the state meet transparency and bias‑mitigation standards. A separate law holds lawyers accountable for core legal work performed by AI, insisting that attorneys retain ultimate responsibility. Additional statutes criminalize the use of deep‑fake technology for impersonation and require gene‑synthesis companies to screen customers for potential misuse.

Human‑review and documentation duties

Across the worker measures, the central theme is documentation. Employers must keep records that show when an AI system was consulted, what data it processed, the rationale for any recommendation, and the human decision that followed. These logs must be retained for a period defined by the state labor department and be made available during inspections or disputes. The requirement for human review does not preclude the use of AI for efficiency; rather, it creates a safety net that ensures a person can intervene, correct or reject an algorithmic recommendation.

Implications for employers

Effective immediately, California employers must audit their existing AI tools to determine whether any are used in disciplinary, termination, layoff or surveillance contexts. Companies will need to establish formal human‑review committees, update policies to include disclosure procedures for AI‑driven workforce actions, and reconfigure surveillance systems to exclude private areas. Failure to comply could trigger civil penalties and enforcement actions by the state labor agency. In practice, organizations can expect a period of adjustment as they integrate human oversight, revise documentation practices, and train managers on the new legal obligations.