Policy

California bill would ban AI emotion recognition at work

AB 1883 prohibits employers from using artificial intelligence to infer workers' emotional states or collect neural data, with civil penalties up to $500 per violation.

Omega Editorial· September 11, 2026· 3 min read

California lawmakers have passed legislation that would prohibit employers from using artificial intelligence systems to detect workers' emotions or collect neural data in the workplace.

Assembly Bill 1883 passed the state Assembly 56-8 on August 30 and now awaits Governor Gavin Newsom's signature. He has until September 30 to sign or veto the measure.

What the legislation prohibits

If signed into law, AB 1883 would amend California's Labor Code to bar employers from deploying workplace surveillance tools that recognize or predict an individual's emotional state. The bill also prohibits collection of neural data, which it defines as information generated by measuring activity in an employee's central or peripheral nervous system.

The legislation applies to what it terms "workplace surveillance tools"—any system, application, instrument, or device that collects employee data, activities, communications, actions, biometrics, or behaviors through means other than direct human observation. This definition encompasses video and audio surveillance, continuous time-tracking tools, geolocation systems, electromagnetic tracking, and photo-optical systems.

Danielle Ochs, a shareholder at Ogletree Deakins in San Francisco, told HR Executive that the bill responds to growing employer adoption of AI-driven monitoring tools. "It seems like almost weekly, employers are introduced to new AI-driven workplace tools with ever-expanding capabilities," Ochs said.

Why it matters

The legislation reflects mounting concern about workplace surveillance overreach as AI capabilities expand. Employee privacy around AI-powered tools has emerged as workers' top concern about artificial intelligence in the workplace, according to a 2026 Prudential Financial study of 3,096 US employees—ranking ahead of accuracy and reliability. California's approach could set a template for other states and influence how employers nationwide think about the boundaries of acceptable workforce monitoring.

Narrow scope after industry pushback

The current version of AB 1883 is deliberately narrow. Earlier drafts drew opposition from employers and industry groups concerned that broader language would inadvertently capture facial recognition technology, standard security cameras, and AI-powered safety systems. Ochs cited tools used to detect distracted or fatigued driving in commercial vehicles as one example opponents sought to protect.

The legislation was tightened in response. Employers can continue using monitoring tools for safety and operational purposes, provided those tools do not engage in the prohibited conduct.

Enforcement and penalties

The California Labor Commissioner and public prosecutors would enforce the measure. Civil penalties of up to $500 per violation would apply. Ochs noted that the bill's language may allow for representative claims under California's Private Attorneys General Act (PAGA), which permits eligible employees to pursue Labor Code-based penalty claims on behalf of the state.

Broader regulatory context

Newsom's track record on AI legislation provides relevant context. He vetoed SB 1047, a broad AI safety bill, in September 2024, arguing it failed to distinguish between high-risk and routine AI applications. However, he signed 17 other AI-related bills that same year and in August signed two bills establishing first-in-the-nation standards for third-party AI audits.

AB 1883's narrowly scoped prohibitions align with the approach Newsom has favored. The European Union's AI Act, whose ban on AI emotion detection in workplaces and education settings took effect in February 2025, prohibits AI systems intended to detect emotional states in those settings except for medical or safety reasons.

Any tools an employer currently uses for workforce analytics, productivity management, or employee engagement monitoring that incorporate sentiment detection, affect recognition, or biometric inference would fall within the bill's scope if signed.

These details were first reported by Insurance Business Magazine.

#workplace ai#emotion recognition#employee privacy#california legislation#workplace surveillance#neural data

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

Sanders Bill Would Ban AI Superintelligence Development

Proposed legislation treats advanced AI creation like nuclear weapons proliferation, with penalties to match.

Via AI Watch · Sep 11, 2026
Policy· 4 min read

AI Researchers Quit Over Extinction Fears They Can't Stop

Insiders say their own models could kill humanity, yet competitive pressure and research needs keep labs racing forward.

Via AI Watch · Sep 11, 2026
Policy· 2 min read

Anthropic Researchers Warn AI Could Cause Human Extinction

Resignations and public warnings from safety researchers at the AI company are fueling U.S. calls for new regulation.

Via AI Watch · Sep 11, 2026