California will require employers to keep a human in the loop before disciplining or firing workers based on automated systems, under a bill Governor Gavin Newsom signed into law on September 30, 2026. Senate Bill 947, authored by state Senator Jerry McNerney, takes effect July 1, 2027, and bars companies from relying solely on an automated decision system, or ADS, to discipline or terminate an employee.
The law, chaptered as Chapter 859 of the 2026 statutes, adds a new section to the California Labor Code governing "Automated Decision Systems in the Workplace." It defines an ADS broadly as any machine-learning, statistical, or AI-driven process that produces a score, classification or recommendation used to assist or replace human judgment and that materially affects workers, while exempting basic tools such as spam filters, firewalls and calculators.
Under the measure, an employer that "primarily relies" on an automated system's output to discipline or fire a worker must have a human corroborate that decision using supporting evidence — which can include supervisor evaluations, personnel files, work product, peer reviews or customer reviews — and must notify the employee in writing that software played a role. If the output cannot be corroborated, or a human reviewer finds it inaccurate or misleading, it cannot be used to justify the decision. Workers can also request a plain-language description of the data used about them.
Enforcement falls to the state Labor Commissioner, which can investigate complaints, issue citations and seek injunctive relief, with civil penalties set at $500 per violation. Public prosecutors may also bring enforcement actions, and affected workers can sue, with the law shifting the burden of proof: once a worker shows an automated system was used, the employer must prove it either did not primarily rely on the system or that it followed the corroboration and notice requirements. The law also forbids retaliation against workers who exercise these rights or file complaints.
SB 947 is a narrower successor to McNerney's earlier "No Robo Bosses Act," SB 7, which Newsom vetoed in October 2025 after concluding its restrictions were too broad and could discourage employers from using technology to reward strong performers based on measures like customer ratings. The new bill dropped a requirement for advance notice before deploying such systems, keeping only a post-decision notice, and explicitly allows online customer reviews as corroborating evidence.
The bill drew support from labor unions, including SEIU California and the California Teachers Association, along with digital rights groups such as the Electronic Frontier Foundation. Opposition came from business groups including the California Chamber of Commerce and technology trade associations such as TechNet and the Business Software Alliance, as well as gig-economy companies Uber, Lyft and DoorDash.
How California's approach compares
The law differs from employment-AI measures advancing elsewhere. Colorado's revised AI law, enacted in May 2026, places documentation obligations on both AI developers and the businesses that deploy their systems, with enforcement by the state attorney general. Connecticut's Senate passed a bill in April 2026 requiring notice and explanations for AI-assisted hiring decisions. California's law, by contrast, is housed in the Labor Code, places all compliance duties on employers rather than software vendors, and is enforced by a labor regulator rather than a consumer-protection agency.
The law's reach into hiring remains limited; its corroboration and notice requirements apply only to disciplinary and termination decisions, not to recruitment or hiring tools. Separately, federal regulators have signaled interest in preempting state AI rules deemed overly restrictive, and it remains unclear whether a labor-focused statute like SB 947 would draw similar scrutiny.
Sources: PPC Land