California Enacts Landmark Law Banning Employers From Using AI Alone to Fire Workers

Labor advocates argue that automated management systems strip human dignity from the workplace, turning careers into data points to be optimized and discarded by algorithms. Business groups counter that restricting these software tools limits productivity and creates immediate legal liabilities over undefined regulatory thresholds.[1][3]

On Wednesday, California Governor Gavin Newsom settled the immediate legislative fight by signing Senate Bill 947. The legislation, known as the "No Robo Bosses Act of 2026," makes California the first state to ban employers from using artificial intelligence as the sole basis for firing or disciplining workers.[1][2][3][5]

Taking effect on July 1, 2027, the law mandates that any termination or disciplinary action driven by an automated decision-making system must be corroborated by a human reviewer. Employers will be required to cross-reference the software's output with traditional metrics, such as managerial evaluations, peer reviews, or personnel files.[1][3][5]

"We need to put a floor and put a marker out there that AI systems really shouldn't be controlling human society," said State Senator Jerry McNerney, the bill's author. He noted that the legislation establishes a necessary boundary for corporate deployment of the technology.[3]

"Nobody wants to be fired by a machine," McNerney added in a separate statement. He emphasized that while algorithmic tools have the potential to increase corporate productivity, they remain highly prone to errors, bias, and misjudgments when left unmonitored.[3][5]

Under the new framework, workers subject to algorithmic discipline must be notified in writing that artificial intelligence played a primary role in the decision. They must also be informed of the specific personal data the system analyzed and be provided with the name of a human contact who can explain the outcome.[1][5]

The scale of algorithmic management

The legislation arrives as corporate reliance on automated management software accelerates across multiple industries. According to the California State Senate, there are currently more than 550 "bossware" products available to employers to monitor productivity and manage workforces.[5]

Lawmakers pointed to the unreliability of current models as a primary justification for the ban. A large-scale study cited by the bill's sponsors found that artificial intelligence assistants misrepresent factual content 45 percent of the time, highlighting the risk of automated errors in employment decisions.[5]

The successful enactment follows a previous legislative failure on the same issue. Governor Newsom vetoed an earlier iteration of the bill in 2025, citing advance-notice requirements that he considered too broad for practical implementation by state employers.[1][3]

Narrowing the scope for passage

To secure the governor's signature this year, McNerney stripped the advance-notice provision from the final text. He also removed language that would have extended the protections to gig workers, a concession that left a significant portion of California's workforce outside the law's umbrella.[1]

Rideshare drivers, who remain classified as independent contractors under Proposition 22, will not be shielded by the new requirements. Labor advocates note that these workers are already subject to the exact automated management practices the bill seeks to curb.[3]

"An algorithm decides what work we get, what we are paid, how our performance is judged, and too often whether we can work at all," said Nicole Moore, president of Rideshare Drivers United. She argued that drivers have lived under algorithmic management long before traditional employees encountered the software.[3]

Despite the concessions, California business groups continued to oppose the final text through the end of the session. Industry representatives argued that the phrase "primarily relies"—the legal threshold that triggers the human-review requirement—remains undefined in the statute, potentially exposing employers to widespread litigation.[1]

A broader workplace privacy package

The "No Robo Bosses Act" was signed alongside a broader suite of labor-backed legislation aimed at establishing guardrails around workplace surveillance. The package positions California as the primary regulator of artificial intelligence in the absence of comprehensive federal action.[2][3]

Among the companion measures is Assembly Bill 1331, which explicitly bans employers from deploying surveillance tools in workplace restrooms. Another measure, Assembly Bill 1883, restricts the use of artificial intelligence to infer workers' emotional states or collect their neural data.[3]

Healthcare workers secured specific protections under Assembly Bill 1979, which bars medical facilities from using artificial intelligence to replace the clinical judgment of licensed professionals. The law mandates that automated systems remain strictly advisory in healthcare settings and extends medical confidentiality to health chatbots.[3]

"We stood up and said that A.I. must support, not replace, the clinical judgement of the nurses and other health care professionals who care for our patients," said Sandy Reding, president of the California Nurses Association.[3]

Newsom also signed Senate Bill 951, which imposes new transparency requirements on corporate restructuring. The law requires employers to provide a 60-day notice when artificial intelligence causes a mass layoff that displaces 25 percent or more of a workforce.[3]

Diverging from federal policy

California's aggressive regulatory posture stands in sharp contrast to recent developments in Washington. The federal government has largely deferred to Silicon Valley on artificial intelligence oversight, with President Donald Trump recently announcing a voluntary accord for tech leaders to "self-police" development.[2][3]

Governor Newsom used the signing event to criticize the federal approach, targeting the reliance on industry self-regulation. He told reporters that the voluntary federal accord is a dynamic that "should scare the hell out of everybody."[2][3]

The contrast was highlighted earlier this year when the short-lived federal Department of Government Efficiency explicitly utilized algorithmic tools to manage personnel. The initiative used the software to identify and terminate probationary government employees via identical template letters generated without individual review.[3]

While Illinois enacted a law in January mandating that employers disclose the use of artificial intelligence in certain hiring contexts, California is the first jurisdiction to outright ban the software from acting as the sole decision-maker.[1]

Enforcement of the new California standards will fall to the state labor commissioner, the attorney general, or local prosecutors. Employers have until the summer of 2027 to audit their management software and establish the required human oversight protocols.[1][5]