California Bans AI Tools That Read Employees’ Emotions or Nervous System Data: FAQs for Employers

California Bans AI Tools That Read Employees’ Emotions or Nervous System Data: FAQs for Employers

California employers will soon be barred from using artificial intelligence to figure out how their employees are feeling. Governor Gavin Newsom signed AB 1883 into law late yesterday, making it illegal for employers to use AI-powered workplace surveillance tools to recognize, infer, or predict an employee’s emotional state, or to collect “neural data” generated by an employee’s nervous system. Below are answers to key questions about what the new law says and what you should do before it takes effect on January 1, 2027.

What does the new law prohibit?

Under the new law, an employer cannot use a “workplace surveillance tool” that uses “AI” on employees for either of two purposes:

• Recognizing, or making inferences or predictions about, an individual’s emotional state
• Collecting neural data

How is “workplace surveillance tool” defined?

A “workplace surveillance tool” is any system, application, instrument, or device that collects or facilitates the collection of employee data, activities, communications, actions, biometrics, or behaviors by means other than direct observation by a person. The statute lists video and audio surveillance, continuous incremental time-tracking tools, geolocation, electromagnetic tracking, photoelectronic tracking, and photo-optical systems as examples, but the list is not exhaustive.

Does the ban cover tools that analyze written communications?
It appears so. Since the definition of “workplace surveillance tool” expressly includes “communications,” software that infers mood or stress from email, chat messages, or call transcripts is likely covered.

How is “artificial intelligence” defined?

The ban applies only to tools that use AI, which the law defines as an engineered or machine-based system that can infer from the input it receives how to generate outputs that influence physical or virtual environments. In practice, that describes most modern analytics products.

How is “neural data” defined?

The law defines “neural data” as information generated by measuring the activity of an employee’s central or peripheral nervous system, and that is not inferred from nonneural information. The clearest examples are EEG headbands and caps that monitor fatigue or attention, and consumer “neurotech” earbuds that claim to track focus. Also worth noting: the ban applies to collecting neural data with an AI-enabled tool, even if you never use that data to make a decision.

What is excluded from the neural data definition?

The law excludes information inferred from nonneural sources, so it likely doesn’t include data derived from things like step counts or general heart rate readings. Harder cases involve wearables that measure skin conductance or muscle electrical activity, which could directly measure peripheral nervous system activity. Keep in mind that a device falling outside the neural data definition may still be prohibited if it uses AI to infer stress, mood, or other emotional states from those readings.

Is there an exception for safety monitoring?

The law “does not prohibit an employer from using a workplace surveillance tool to ensure safety, or otherwise if it does not meet the specifications” of the ban.

How broad is the safety exception?

This is an open question. Some early analyses of the bill have read this exception narrowly, meaning that safety monitoring tools would remain lawful only if they do not engage in the prohibited conduct.

But the text of the law supports a broader reading. The phrase “or otherwise if it does not meet the specifications” suggests that tools used “to ensure safety” are protected even if they do meet those specifications. Under that interpretation, tools like a fatigue-monitoring headset on a heavy equipment operator could be lawful despite collecting neural data.

Until state officials or a court address the question, you should work with your FP attorney and build your monitoring program based on your specific circumstances and risk tolerance.

Who is covered?

The law applies to private employers of every size and to public employers, including all branches of state government, the University of California, the California State University, cities and counties (including charter cities), and school, community college, special, and transit districts. It also reaches labor contractors, farm labor contractors, foreign labor contractors, and businesses that contract with labor contractors, including client employers and entities that contract for property services.

Who does the law protect?

The law protects “employees.”

• Earlier versions covered independent contractors, but that language did not survive.
• The statute also does not address job applicants on its face, although current employees who apply for internal positions remain covered.

Are any industries exempt?

There is an exemption for employers developing aircraft for the national airspace or products and services for national security, military, space, or defense purposes. But the exemption only applies to operations covered by a federal statute, federal regulation, or binding federal contract where the tool is reasonably necessary for compliance.

How will the law be enforced?

The Labor Commissioner may investigate alleged violations, order temporary relief while an investigation is pending, issue citations, and file a civil action. Public prosecutors may also enforce the law. In a civil action brought by either, the petitioner may seek injunctive relief, punitive damages, and reasonable attorney’s fees and costs.

What are the penalties?

Employers can face a penalty of up to $500 for each violation. The statute does not define what counts as a single violation, so exposure could multiply quickly if each affected employee or each day of use is counted separately.

Can employees sue us if they believe the law has been violated?

The final bill removed the private right of action that appeared in earlier drafts. But Labor Code Section 1582(d)(2) refers to “an employee” recovering a penalty “as a statutory penalty paid to the employee.” Expect plaintiffs’ attorneys to point to that section and argue that the law’s civil penalties can be pursued through representative claims under the Private Attorneys General Act (PAGA).

What should we do to prepare for the new law taking effect?

• Inventory wearables and safety technology. Identify any fatigue monitors, EEG devices, or biometric wearables in use. For each, determine whether it collects neural data or infers emotional states, and whether it serves a documented safety purpose. For any tool you keep in use under the safety exception, document the specific safety purpose and why you need the tool to achieve it.

• Inventory AI features, too. Review your contact center, collaboration, video meeting, HR analytics, and employee wellbeing platforms for anything labeled sentiment, tone, engagement, mood, stress, wellbeing, or burnout.
• Send vendors written questions. Ask each vendor whether its product infers or predicts the emotional state of any individual, including from text, voice, or video; whether it collects data measuring nervous system activity; and whether those features can be disabled for your California employees. Get the answers in writing and consider contract amendments that require notice before new features of this kind are turned on.
• Decide whether to switch features off for California or everywhere. Multistate employers will need to choose between a California-only configuration and a company-wide change. If you have operations in the European Union, the EU AI Act has prohibited AI emotion recognition in the workplace since February 2025 (with exceptions for medical and safety reasons), so coordinating your global approach may be simpler than maintaining separate settings. Keep in mind that the final law focuses on whether a tool is used for the prohibited purposes rather than what it is capable of (an earlier version reached tools “capable of” those functions), so disabling these features and documenting that they are turned off can be a meaningful compliance step.
• Train managers and HR on what they cannot use. Make clear that AI-generated inferences about how an employee feels are off limits, whether they come from a camera, a microphone, or a chat log.
• Update monitoring policies and notices. Revise employee monitoring policies, handbooks, and notices so they no longer describe emotion-inference or neural data features you have turned off, and so they accurately reflect the tools that remain in use.
• Don’t forget the CCPA. If your business is subject to the CCPA, neural data is considered “sensitive personal information,” and monitoring tools that remain in use may trigger notice at collection, right-to-limit, and risk assessment obligations.
• Document your compliance steps now. Under the 2024 PAGA reforms, employers that take reasonable steps to comply before receiving a PAGA notice may be eligible for reduced penalties. Keeping records of your tool inventory, vendor responses, and configuration changes can help support that position.

Conclusion

We will continue to monitor developments and provide updates as warranted. Make sure you are subscribed to Fisher Phillips’ Insight System to get the most up-to-date information direct to your inbox. If you have questions, contact your Fisher Phillips attorney, the authors of this Insight, or any attorney in our California offices, our Privacy and Cyber Practice Group, or our AI, Data, and Analytics Practice Group.