What happened
Law and the Workplace reported that California enacted three laws regulating workplace AI, with provisions taking effect in 2027.
Law and the Workplace, a Proskauer Rose labor and employment blog, reported that California Gov. Gavin Newsom signed three bills at the end of the 2026 legislative session. The report says SB 947, effective July 1, 2027, requires human corroboration when an employer primarily relies on an automated decision system’s output to discipline or terminate an employee. If the output cannot be corroborated or is inaccurate, incomplete, or misleading, the employer cannot use it for that decision. Affected employees must receive written notice of the AI reliance, human corroboration, and a contact for more information, and may request an objective description of their data used by the system.
The report says SB 947 also bars automated decision systems from inferring protected status or predicting and penalizing employees for exercising legal rights. AB 1883, effective January 1, 2027, prohibits AI-powered workplace surveillance from recognizing, inferring, or predicting emotional states and from collecting neural data, while creating an express safety-related exception whose scope remains unresolved. SB 951 adds technology-displacement disclosures to certain California WARN notices beginning January 1, 2027, including affected job functions, locations, occupations, employee counts, and the type of technology involved.
Source details: lawandtheworkplace.com ↗
Why it matters
The laws could affect how California employers use automated systems in discipline, employee surveillance, and workforce planning. Workers may receive more information about AI-assisted decisions and automation-driven layoffs, while employers may need to document human review and audit surveillance tools. The supplied report does not establish the laws’ enforcement record, penalties, implementation guidance, or practical effect on employers and employees.
These measures place specific procedural and disclosure obligations around AI systems that can directly affect employment. The practical implication for employers is that automated outputs may need an auditable human review before certain disciplinary or termination decisions, while workforce reductions attributed substantially to automation may require more detailed public-facing notices. For workers, the laws could provide more visibility into AI-assisted decisions and technology-driven layoffs. These effects are reported by Law and the Workplace and have not been independently confirmed here from the statutory texts or government guidance.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
Why can ethical evaluation not be reduced to one model score?
What to watch next
Key questions include how regulators and courts interpret the workplace-safety exception for surveillance tools, how employers define primary reliance on an automated decision system, and what summaries the Employment Development Department publishes. The report also leaves enforcement mechanisms, penalties, and implementation guidance unknown.
The report identifies the safety exception in AB 1883 as an important unresolved issue. Future interpretation will determine which workplace monitoring uses remain permissible when an employer claims a safety purpose.
The supplied source does not state the laws’ penalties, enforcement procedures, regulatory guidance, definitions beyond its summary, or how the requirements will apply to particular tools and workplaces. It also does not provide evidence about how many employers currently use systems covered by the laws.