California Gov. Gavin Newsom signed SB 947, the No Robo Bosses Act, barring employers from relying solely on artificial intelligence to fire or discipline workers and restricting AI as the principal basis for those decisions. Employees must receive notice when an employer primarily relies on AI, a description of the data used, and a human contact who can explain the decision.
What Workers Won
“When working people organize, we get results,” Lorena Gonzalez, president of the California Federation of Labor Unions, AFL-CIO and lead sponsor of the act, said after Newsom signed it. Workers across California had demanded state regulation of AI in workplaces, she said. Organized labor pushed for limits after employers began using AI in decisions that could harm workers.
Under SB 947, employers that primarily rely on AI output must have a human reviewer corroborate termination or disciplinary decisions with additional information. The law names managerial evaluations, peer reviews and personnel files as examples. It doesn’t prohibit employers from using AI in those decisions. But affected employees must receive a human review and written disclosure.
“No worker should ever be fired or disciplined by a machine, AI or not,” California State Senator Jerry McNerney, the bill’s author, told CNBC. He said AI systems can boost productivity but have also made errors and misjudgments and exhibited bias. “AI must remain a tool controlled by humans, not the other way around.” McNerney, a Democrat, first introduced the act in 2025.
The Electronic Frontier Foundation called the law a “strong step toward giving workers the protections they need in workplaces that use automated decisionmaking systems.” The nonprofit said it would continue working with labor groups on policies to protect human dignity in the workplace.
Where the Bill Was Narrowed
Newsom vetoed an earlier version in October, even though both chambers of the state Legislature had passed it with overwhelming majority support. He objected to a requirement that businesses notify workers in advance whenever AI could affect work conditions. Newsom said unregulated use of automated decision-making software could harm workers, but argued the bill imposed “unfocused notification requirements” on businesses using even “the most innocuous tools.”
When McNerney reintroduced the bill in February, he removed the advance-notice requirement and language extending protections to gig workers. Rideshare companies Uber and Lyft had criticized that language. Business leaders continued to oppose SB 947. Robert Singleton, the Chamber of Progress’ senior director of policy and public affairs for California and US West, urged Newsom to veto it, arguing that “primarily relies” was undefined and left employers without an objective standard. He warned that uncertainty could discourage technology use for consistency, identifying safety risks and helping managers make informed decisions.
CNBC described SB 947 as the first law of its kind to become law. Illinois’ Public Act 103-0804 requires notice when employers use AI for specified employment purposes, but doesn’t prohibit agentic digital systems from making employment decisions. A federal bill with the same name, introduced in June by U.S. Senators Ed Markey (D.-Mass.) and Brian Schatz (D.-Hawaii), would bar employers from relying on automated decision systems for work-related decisions; CNBC said it was effectively dead on arrival. Similar measures proposed in New York, Louisiana and New Jersey hadn’t advanced.
AI at Work
An OECD survey published late the previous year found 90% of U.S. managers said their firms had adopted at least one tool to “instruct, monitor or evaluate workers,” the highest share among countries surveyed. Former employees suing Meta alleged in a lawsuit filed in July that AI-assisted systems ranked and selected workers for layoffs, disproportionately affecting employees who had taken medical or family leave. Meta denied the allegations.
A May survey by United for Respect said Walmart and Amazon workers were increasingly concerned about automated human-resources decisions. The group failed last summer to persuade Walmart shareholders to require management to disclose more about AI use with its workforce.
Newsom also signed a broader executive order earlier in September addressing potential existential risks from AI models and a measure establishing a state framework for independent evaluation and auditing of AI models. He said the federal government was abdicating “its responsibility to protect Americans.” The Washington Post reported that Newsom signed laws on Wednesday aimed at protecting workers from AI threats, including potential job losses and workplace surveillance.