California now wants employers to say when an AI caused the layoffs. The law has a big gap.
Gov. Newsom signed two worker bills on September 30. One makes companies label AI-driven layoffs. The other bars firing someone on an algorithm's say-so alone. Here's what the text actually requires.

On September 30, Gov. Gavin Newsom signed 13 AI-related bills in one batch, according to his office’s list. Most of the headlines went to a tongue-in-cheek executive order his office described as “permanently declaring” that AI be called “Artificial Intelligence” in California. The two bills that matter are about work, so I read the chaptered text of both.
What does SB 951 make employers do?
California already has a mini version of the federal WARN Act. If a company with 75 or more employees lays off 50 or more people at one site within 30 days, or closes or relocates it, it has to give 60 days’ written notice to the workers, the state’s Employment Development Department and local officials.
SB 951 adds a paragraph to that notice, but only when a layoff is “caused in whole or in substantial part” by AI or other automated technology replacing positions. In that case the notice must give the number and type of jobs affected, the job functions being automated, and the category of AI system responsible. It also has to open with the sentence “This notice is for a technology displacement.”
The state then has to publish a summary of those notices and a quarterly statewide tally. Separately, the EDD owes the Legislature a report by January 1, 2028 on how AI is changing hiring, including a look at whether employers should have to announce when they simply stop hiring for a role because of AI.
What does SB 947 do?
It covers automated decision systems used on employees. Starting July 1, 2027, an employer can’t rely solely on such a system to discipline or fire someone. If it leans on one primarily, a person has to check the call against other evidence, such as a manager’s review, personnel files or witness interviews. If the human can’t back up the system’s output, the employer can’t use it.
Workers also get two things. They can ask for a plain description of their own data that the system used, and they must get a written notice, separate from the other paperwork, saying a system drove the decision and naming a human to contact. The Labor Commissioner enforces it, with a $500 civil penalty per violation.
Where are the gaps?
Three, as far as I can tell from the text.
First, the layoff rule only triggers on mass layoffs at a covered site. Fifty people in 30 days. A company that quietly lets 20 people go every month, at several sites, never files this notice at all.
Second, the “substantial part” test is the employer’s call to make. The bill requires the label when AI substantially caused the layoff, but nothing I read says who decides that or how it’s checked. A company can say the cuts were about “restructuring” and the AI part may never show up on the form.
Third, SB 947 doesn’t apply to workers covered by a union contract that explicitly waives it. The text also exempts some aircraft and defense work. I haven’t found the effective date for SB 951, so I’m not going to guess one.
What did the unions say?
The California Federation of Labor Unions called the bills “first-in-the-nation,” according to Newsom’s office. I’d want to hear from employers’ groups too. I didn’t find a statement from them in the coverage I could read, so I can’t tell you how business is taking it.
My read
I like the label more than I expected to. Right now there’s no official count of jobs lost to AI anywhere in the US, only company statements and guesses. A state database of notices would be thin, since the 50-person threshold filters out most cases, but it’d be real filings, with the job functions named. That’s more than we have.
I’m less sure the “human must check” rule in SB 947 changes much. A manager who clicks “agree” on a score takes ten seconds. The better test is whether the worker’s right to see their own data gets used, and whether the first Labor Commissioner cases show up before the 2027 start date.
Related: the White House AI accord and the bill in Congress with million-dollar fines covers the federal side, where nothing like this exists yet.
Sources
- Office of Gov. Gavin Newsom: California’s nation-leading AI framework just got stronger, September 30, 2026
- California Legislative Information: SB 951, Chapter 860 (chaptered text)
- California Legislative Information: SB 947, Chapter 859 (chaptered text)
- Transparency Coalition: Gov. Newsom wraps California term by enacting 11 more laws on AI safety, September 30, 2026
- Transparency Coalition: AI Legislative Update, October 2, 2026