The new California AI laws signed on 30 September 2026 are the most far-reaching set of artificial intelligence rules any US state has passed. Thirteen bills landed at once, and together they decide whether an algorithm can fire you, watch you, impersonate you or overrule your doctor.
California sets the national baseline in practice, because almost no company builds one HR system for California and another for everywhere else. If you work in the US, these California AI laws will probably reach you.
Here is what each of the new California AI laws actually does, in plain English.
The Headline Act: SB 947, the “No Robo Bosses Act”
SB 947, authored by Senator Jerry McNerney, is the bill everyone is talking about. It prohibits employers from relying only on an automated system when making a disciplinary or termination decision.
In other words, a human has to be in the loop before you lose your job.
What it does not do is ban AI from the process. An employer can still use software to flag performance, score productivity or rank candidates. It simply cannot let the software press the button by itself.
That distinction matters. These California AI laws regulate the final decision, not the analysis that leads up to it.
Who the new California AI laws affect most
- Warehouse, delivery and gig workers managed by productivity scoring systems.
- Call centre and support staff graded by automated quality monitoring.
- Anyone whose performance review pulls from an algorithmic dashboard.
SB 951: California AI Laws Now Require Layoff Disclosure
Senator Eloise Gómez Reyes’ SB 951 is the quieter bill with the bigger long-term effect. It requires employers to disclose when mass layoffs, relocations or terminations are the result of AI systems.
Until now, “restructuring” could quietly mean “we automated it.” This forces the label on.
For economists and journalists, this is the first real dataset on AI-driven job displacement in the United States. For workers, it is the difference between a vague explanation and a documented one.
The Surveillance Bills: AB 1331 and AB 1883
Two of the California AI laws target monitoring rather than decisions.
AB 1331 (Assemblymember Sade Elhawary) bans workplace surveillance tools in bathrooms. That such a law was necessary tells you where things had got to.
AB 1883 (Assemblymember Isaac Bryan) protects employee privacy more broadly from AI-powered workplace surveillance.
Together they draw a line that has been blurry since remote work normalised keystroke logging, webcam checks and location tracking.
Healthcare: AB 1979 and SB 503
Two bills keep clinical judgement with clinicians.
AB 1979 (Assemblymember Mia Bonta) ensures doctors can exercise professional judgement when AI clinical decision tools are involved, and requires bias reduction in those systems.
SB 503 (Senator Akilah Weber Pierson, a physician) reinforces physician authority in AI-assisted healthcare decisions.
The practical target here is insurer-side automation: algorithms that deny coverage or recommend against treatment without a doctor signing off.
SB 574: Lawyers Cannot Outsource Their Brains
Senator Thomas Umberg’s bill prohibits lawyers from fully delegating core legal work, such as drafting briefs, to AI.
After a string of hallucinated-citation scandals in courts nationwide, this one was inevitable. Among the California AI laws it is the clearest signal that professional accountability cannot be subcontracted to a model.
Deepfakes and Transparency: SB 1111, AB 2713 and SB 1000
SB 1111 (Senator Angelique Ashby) updates civil and criminal law to protect likeness rights and prevent false impersonation through deepfakes.
AB 2713 (Assemblymember Buffy Wicks) strengthens the California AI Transparency Act by prohibiting the deletion of digital watermarks and allowing metadata access to identify AI-generated content.
SB 1000 (Senator Josh Becker) extends those transparency requirements to keep pace with newer systems.
The watermark provision is the technically interesting one. Stripping provenance data becomes a legal act, not just a technical one, which changes the incentives for every tool that touches generated media.
The Remaining Bills
- AB 1864 (Marc Berman): gene synthesis companies must verify customers and screen genetic material, closing a biosecurity gap around pathogens such as polio and Ebola.
- AB 2392 (Mike Fong): creates generative AI training and procurement requirements for students, faculty and staff across public higher education.
- SB 1159 (Christopher Cabaldon): ensures public comments at local government meetings come from real people, not AI bots.
What the California AI Laws Mean for Employers
If you run or advise a business, the compliance work is mostly procedural rather than technical.
- Document the human. Name who reviews automated recommendations before adverse actions, and keep the record.
- Inventory your tools. Most companies underestimate how many systems already score employees.
- Fix your layoff notices. If automation drove the decision, SB 951 means saying so.
- Audit vendors. Your HR software provider’s claims are now your legal exposure.
SB 947 carries an effective date of 1 July 2027, which gives organisations roughly twenty-one months. That is less time than it sounds for anyone replacing an entrenched HR stack.
What It Means for Workers
Three practical takeaways from the California AI laws:
Ask how decisions are made. You are increasingly entitled to know whether a system scored you and who reviewed it.
Keep your own records. If a dispute arises, your documentation of performance and communication is what a human reviewer will weigh.
Read layoff paperwork properly. The disclosure requirement only helps if someone actually reads the disclosure.
California AI Laws FAQ
Can AI legally fire someone now?
Not on its own, once SB 947 takes effect. A human must review and make the final call.
Do these laws apply outside California?
Formally, no. Practically, large employers tend to standardise policy nationally rather than run parallel systems.
When do the California AI laws take effect?
Dates vary by bill. SB 947 is set for 1 July 2027; others phase in earlier. Check each bill individually.
Is this the strictest AI regulation in the US?
At state level, yes, and by a wide margin. Colorado and Illinois have narrower AI employment rules; California’s package is the broadest.
The Bottom Line
The California AI laws do not slow AI adoption. They attach a name to it. Every consequential decision now needs a human who can be asked why, which is a modest requirement that will nonetheless reshape how automated management tools are built and sold.
Expect other states to copy the California AI laws package in their 2027 sessions, and expect the federal debate to use it as the reference point.
This article is general information, not legal advice. Consult an employment attorney about your specific situation.
For more on where the underlying technology is heading, see our guides to the best AI agents of 2026 and the economics behind the AI boom.
Spending your evenings reading legislation? You have earned better downtime. KenoIPTV delivers thousands of live channels, sports and on-demand titles in one app at an unbeatable price. Explore KenoIPTV plans here.
Bill details via the Office of Governor Gavin Newsom and the National Law Review.
