California lawmakers want to curb the use of "emotional AI" in the workplace, a move that would place the state at the forefront of regulating emotion-recognition technology in employment settings.

The legislative push targets tools that infer workers' emotions, moods or mental states from facial expressions, voice patterns, keystrokes or other behavioral signals. For HR teams, the proposal touches directly on decisions about hiring assessments, productivity monitoring, wellness programs and call-center analytics — any people process that relies on algorithmic inference about how an employee feels.

The term "emotional AI" covers a broad class of systems: software that scores job candidates' video interviews for enthusiasm, platforms that flag "disengaged" employees based on tone of voice, and surveillance tools that read facial cues during the workday. California's action signals that lawmakers increasingly view these inferences as sensitive personal data — and potentially as a basis for discriminatory or unfair employment decisions.

What it means for employers

Companies operating in California should expect scrutiny of any vendor tool that claims to measure emotion, engagement or sentiment. That includes:

• Hiring technology. Automated video interview platforms that evaluate candidate affect or personality traits could fall within the law's reach.
• Workforce monitoring. Tools that analyze voice tone in customer service calls or facial expression in workplace cameras.
• Wellness and engagement platforms. Systems that infer burnout, stress or morale from behavioral data.

The practical compliance question for HR leaders is inventory: which current systems collect or generate emotion-related inferences, which vendors supply them, and what contractual and notice obligations already apply. Employers that wait for final statutory language before auditing their HR tech stack may face a compressed remediation window.

California's regulatory trajectory

California has repeatedly acted first on workplace technology issues, and its rules often become de facto national standards because multi-state employers align policies to the strictest applicable regime. A restriction on emotional AI would extend a pattern that includes the state's privacy regime, which already grants consumers rights over their personal information and restricts certain data uses.

Globally, the direction is similar. Regulators and lawmakers in other jurisdictions have moved to limit emotion-recognition systems, particularly in workplaces and schools, reflecting concern that the underlying science — inferring internal states from external signals — remains unreliable and can produce biased outcomes across demographic groups.

For HR, the analytical questions are straightforward. Does the tool actually measure what it claims? Can employees contest an emotion-based inference? And can the employer demonstrate that any employment decision resting on such inference is job-related and consistent with business necessity?

Adoption questions for vendors

Employers evaluating emotional AI tools — or already deployed platforms — should press vendors on capability and evidence rather than marketing claims:

1. Validation. What peer-reviewed or third-party evidence supports the tool's accuracy in detecting emotion, and in which populations?

2. Adverse impact. Has the vendor tested the system for disparate performance across age, race, gender or disability?

3. Data handling. What emotional inferences are stored, for how long, and who inside the employer can access them?

4. Decision use. Can the tool's outputs influence hiring, promotion, discipline or termination — and can those decisions be reconstructed without the inference?

5. Sunset planning. If California restricts the technology, can the vendor's product function with emotion features disabled?

That last question may prove decisive. Multi-state employers facing a California restriction will need the ability to toggle off emotion-detection features for California workers — or for everyone, if a single national configuration is simpler to govern.

What to watch

The legislative process will determine scope: whether the restriction applies only to workplace uses, whether it bans the technology outright or requires consent and disclosure, and what enforcement penalties attach. HR and legal teams should track committee activity, amendment language and effective dates as the bill advances.

If California enacts the restriction, expect other states with active workplace privacy agendas to follow, and expect employee plaintiffs' attorneys to test existing claims — discrimination, privacy, wiretapping — against emotion-detecting tools even before new statutory remedies take effect.