Can AI fire you? Right now, in many places, the honest answer is “sort of.” Software already tracks how fast people work, flags “low performers” and suggests who should go. On September 30, 2026, California became the first US state to say clearly: a computer can help, but a human must make the final call.
The new law is called the No Robo Bosses Act (Senate Bill 947). Governor Gavin Newsom signed it after he had blocked an earlier version in 2025. This guide explains what the law does, what it does not do, and what it means if you work outside California.
Can AI fire you under California’s new law?
No — not on its own. Under the No Robo Bosses Act, a California employer cannot use an AI system as the only thing behind a decision to fire or discipline a worker. If the AI played a big part, a human must check the decision against other proof, such as manager reviews or personnel files. The law starts on July 1, 2027.
Here is the key idea in one line: AI can point at a problem, but a person has to confirm it before anyone loses their job.
The law uses the phrase “automated decision system.” That simply means software that makes or shapes a decision about people — for example, a tool that scores workers, ranks them, or flags them as a risk.
What the No Robo Bosses Act actually requires
According to reporting by Bloomberg Law and AI Weekly, the law has three main parts:
- No AI-only firings. An employer cannot rely only on an AI tool to fire or discipline someone.
- A human double-check. If the AI was the main reason for the decision, a person must back it up with other information. The law names examples: manager evaluations, peer reviews and personnel files.
- A written notice. The worker must be told in writing that AI was used. They must also get a description of the data the system looked at, and the name of a human they can contact about the decision.
Think of a warehouse worker whose scanner data says they are “too slow.” Under this law, the software alone cannot end their job. A manager would need to look at the full picture. Maybe the worker was training a new colleague. Maybe their scanner was broken for a week.
Why did Newsom say no in 2025 and yes in 2026?
In October 2025, Newsom vetoed (refused to sign) a bill called SB 7, an earlier version of the same idea. He said it put “overly broad restrictions” on how employers could use these tools.
The 2026 version is narrower. The biggest change: the old bill wanted workers warned before AI was used. The new law drops that. Instead, the worker gets the notice when the decision arrives. That made it easier for businesses to accept.
The law is part of a bigger push. On the same day, Newsom signed 12 other AI bills, according to Gizmodo. They cover topics like AI-driven layoffs, workplace monitoring, deepfakes, medical AI and lawyers using AI.
What the law does not cover
This part matters, because headlines can make the law sound bigger than it is.
- Hiring is not covered. The law is about firing and discipline. Bloomberg Law notes it does not deal with hiring decisions. AI tools that sort job applications are a separate issue.
- AI is still allowed. Companies can keep using AI to track work, score performance and flag problems. They just cannot let it make the final decision alone.
- It is not live yet. The rules start on July 1, 2027. Until then, nothing changes for California workers.
- It is California only. Workers in Texas, New York or anywhere else in the US do not get these rights from this law.
If you are curious how AI is reshaping jobs more widely, our guide on why AI is not one job anymore looks at the other side of the story.
Has AI really fired people before?
Yes. The best-known case is from the Netherlands. In 2021, a court in Amsterdam ordered Uber to reinstate six drivers. The court found they had been fired “based solely on automated processing” — in plain words, a computer decided, and no human really checked. The court also ordered Uber to pay the drivers compensation, according to the UK Human Rights Blog.
Workers’ groups call this “robo-firing.” It happens most in jobs run through apps: ride-hailing, food delivery and warehouse work. In these jobs, an app often does the work a manager used to do.
Can AI fire you if you live outside California?
It depends on where you live. In Europe, a data law called the GDPR already gives people the right not to face major decisions made only by a machine. That is the rule the Amsterdam court used. In many other countries, there is no law written this clearly about AI and firing yet.
Here is a quick look by region:
- Europe and the UK: Article 22 of the GDPR protects people from big decisions made “solely” by automated systems. On top of that, the EU AI Act lists AI used in hiring and managing workers as “high-risk,” which brings extra rules. Those extra rules were delayed: they now start on December 2, 2027, after the EU approved its “Digital Omnibus” package in June 2026.
- United States: There is no national rule like this. California now leads, and other states may copy it.
- India, Australia and elsewhere: General labour and data laws still apply. But we could not find a rule as specific as California’s. If you are worried, ask your employer or a local labour lawyer how decisions about you are made.
The short version: Europe protects workers through data law, California now does it through job law, and most other places have not decided yet.
What you can do if you think an AI judged your work
You do not need to be a tech expert to protect yourself. A few simple steps help almost anywhere:
- Ask directly. “Was any software or AI tool used in this decision?” It is a fair question.
- Ask for the data. In Europe and the UK, you can request the personal data a company holds about you. From July 2027, California workers will get a description of the data used.
- Ask for a human. Request a meeting with a real person who can explain the decision.
- Keep your own records. Save emails, schedules and notes that show the full story — things a tracking tool might miss.
Platform work is where this question hits hardest. Our piece on Amazon Mechanical Turk closing shows how quickly app-based work is changing.
Frequently Asked Questions
Is it legal for AI to fire someone?
In most places, no law bans it outright yet. California’s No Robo Bosses Act will require a human to check AI-driven firings from July 1, 2027. In Europe, the GDPR already protects people from major decisions made only by machines.
When does the No Robo Bosses Act take effect?
The law was signed on September 30, 2026, and starts on July 1, 2027. Until then, California employers do not have to follow its rules.
Does the No Robo Bosses Act cover hiring?
No. The law covers firing and discipline, not hiring. AI tools that screen job applications are not covered by this law.
What must an employer tell a worker under the law?
The worker must get written notice that AI was used in the decision. They must also get a description of the data the system used, and the name of a human contact who can explain it.
Do I have this protection if I live outside California?
Not from this law. People in the EU and UK have some protection under the GDPR. In most other countries, including the rest of the US, there is no rule this specific yet.
