The California State Capitol in Sacramento, where SB 951 added AI layoff disclosures to the state's WARN Act.
← Back to Blog
AI Governance

SB 951 Makes California Count AI Layoffs. The Harder Count Is the Jobs Never Posted.

Head portrait of Alex Goryachev
Alex Goryachev·October 16, 2026·5 min read

From January 1, 2027, California WARN Act notices for AI-driven mass layoffs must open with "This notice is for a technology displacement," and SB 951 also orders the state to study the jobs AI ends before anyone is laid off.

Bring Alex to your organization: keynotes, workshops and advisory, in person or virtual. Check Alex's availability

Starting January 1, 2027, a California employer whose mass layoff is caused "in whole or in substantial part" by artificial intelligence must print one sentence at the top of its WARN Act notice: "This notice is for a technology displacement." That requirement is the heart of SB 951, written by Senator Eloise Gómez Reyes (D-Colton) and signed by Governor Gavin Newsom on September 30. The most revealing line in the new law sits further down, in a section that asks the state to start counting the AI jobs that disappear before anyone receives a notice at all.

Cal-WARN already requires 60 days' written notice to affected workers, the Employment Development Department (EDD), and specified local officials when an establishment with 75 or more employees lays off 50 or more people within 30 days. SB 951 keeps that clock and those thresholds. When AI or automation drives the cut, the notice must add four things: the number of positions, with their occupations and work locations; the job functions that will be automated; the category of technology involved; and that technology-displacement header. EDD will publish summaries of these notices and a quarterly statewide tally.

SB 951 adds no new penalties and leaves Cal-WARN's existing remedies in place. Employment lawyers at Fisher Phillips note that those penalty provisions point to a different subdivision than the one holding the new disclosures, so the consequence of a wrong AI disclosure remains an open question.

The bill changed shape on its way to the governor's desk

As introduced on February 2, SB 951 asked for more: it proposed 90 days' notice and reached AI-related layoffs or hiring reductions touching as few as 25 workers. The signed version folds the disclosure into the 60-day Cal-WARN process employers already follow, and turns the hiring question into a study. That is what careful legislating looks like: one clock for employers to track, and a data request before a mandate.

ProvisionSB 951 as introduced (Feb. 2, 2026)SB 951 as signed (Sept. 30, 2026)
Advance notice90 days60 days (existing Cal-WARN clock)
Trigger size25 workers or 25% of the workforce, whichever is less50 or more layoffs in 30 days at an establishment of 75 or more
AI-driven hiring reductionsCovered by the notice requirementEDD report to the Legislature due January 1, 2028
Public dataQuarterly EDD reports to the LegislatureEDD notice summaries plus a quarterly statewide tally

Senator Reyes framed the bill around pace from the first day:

"The reach and speed at which AI is transforming our economy is astounding." (Senator Eloise Gómez Reyes, introducing SB 951, February 2, 2026)

A layoff has a date

Notice laws were built around events with a date. The federal WARN Act of 1988 grew out of an era of plant closings, when the decision and the last shift both sat on a calendar someone could circle. Cal-WARN inherited that design, and it remains a precise instrument for measuring a layoff.

AI moves through a company on a different schedule. In the executive conversations I sit in, automation tends to show up first as a requisition that stays closed after someone retires, and then another. A team shrinks one unreplaced role at a time, and no single 30-day window ever crosses the 50-person line. Even when a cut is large enough to trigger the notice, the 60 days count back from the layoff order, and in the cases I see, the automation decision was made months earlier, when the purchase order was signed. The notice reports the last step of a choice that began long before.

That is the governing idea behind the AI governance work I follow most closely: rules written for a slower world, applied in good faith to change that arrives as a process. This summer I wrote about layoffs where nobody would say AI did it. For the cuts big enough to trigger Cal-WARN, SB 951 now asks employers to say it in the first line.

The section that counts what never happened

New Labor Code Section 1402.7 defines "technological cessation in hiring" as "the permanent ending of hiring or contracting for a particular occupation or position" because of AI or automation, whether or not anyone currently in that job is let go. EDD must report to the Legislature by January 1, 2028, on AI's effect on hiring, and may recommend a public notice for exactly this. The Legislature deserves credit for writing it: the clause names the mechanism a 1988-era framework was never built to see, and gives EDD the job of building a way to see it.

Consider who that section is for. A layoff notice reaches someone. A job that stops being posted reaches the graduate who would have applied next spring, and the clerk two desks over whose next step up just disappeared. Neither one gets a letter, because there is no address to send it to.

That is the half-life of skills in its most literal form: the market for a skill expires while every person who holds it is still employed. A layoff has a date. A hiring stop only has an absence.

I have watched a scoreboard miss a shift before

I lived through the Napster years from inside the digital music business, at Napster and then Liquid Audio. Napster launched in June 1999. The Billboard Hot 100, the scoreboard the whole industry read, began including digital downloads with its February 12, 2005 issue. The chart counted what it was built to count, sales at a register, while listening moved somewhere no register could see. What I took from those years is simple: when the official count runs behind the behavior, the people closest to the work see the shift first, provided someone asks them.

SB 951 gives California its first recurring public count of AI displacement, published every quarter by EDD. Multiply those tallies across the state where much of this technology is built, and they become a map of where relearning has to happen, occupation by occupation, for the families who live there.

What any company can start counting now

Every employer can begin the hiring count today, more than a year before EDD's report reaches Sacramento. I am taking two questions into my own meetings this quarter; borrow them for yours. Which roles did we stop backfilling this year, and did AI play a part in that choice? And if EDD recommends a hiring-cessation notice in 2028, could we fill one in accurately from the records we keep now?

If you hold one of those roles, the question belongs to you too. Ask whether your job is still being hired for anywhere in your company. The answer will tell you more about the half-life of your skills than any layoff notice will.

California has written the first sentence employers must put on a technology displacement notice. Here is the one I would ask every leadership team to draft for itself: what would the notice say about the jobs you stopped posting? Take that into your next meeting. And if you see it differently, I'm easy to find.

Sources: SB 951 (Chapter 860, Statutes of 2026), enrolled text via LegiScan; Office of Senator Eloise Gómez Reyes, February 2, 2026; Littler, October 2, 2026; Fisher Phillips, October 1, 2026; Ogletree Deakins via National Law Review, October 9, 2026; Forbes, October 5, 2026; Billboard Hot 100 methodology, February 12, 2005 issue.

Does California require notice for AI layoffs?

Yes, starting January 1, 2027. Under SB 951, a Cal-WARN notice for a mass layoff caused in whole or in substantial part by AI must open with "This notice is for a technology displacement" and list the affected positions, the job functions being automated, and the type of technology involved.

When does SB 951 take effect?

SB 951 takes effect January 1, 2027. Governor Gavin Newsom signed it on September 30, 2026, as Chapter 860 of the Statutes of 2026.

How many days notice for a layoff in California?

Cal-WARN requires 60 days' written notice for a mass layoff of 50 or more employees within 30 days at an establishment with 75 or more employees. SB 951 keeps that 60-day period, although the version Senator Eloise Gómez Reyes introduced in February 2026 proposed 90 days for AI-driven layoffs.

What is technological cessation in hiring?

SB 951 defines it as the permanent ending of hiring or contracting for a particular occupation or position because of AI or automation, whether or not current workers lose their jobs. California's Employment Development Department must report on it to the Legislature by January 1, 2028, including whether a public notice should be required.

Here is what makes Alex a credible voice on this topic: He watched an industry's scoreboard fall behind its listeners from inside Napster and Liquid Audio, then spent 20 years at Cisco, where as Managing Director of Innovation Strategy and Head of Global Innovation Centers he shaped a $1.1B innovation portfolio that generated more than $400M in revenue, so he knows how far a technology decision travels before any notice like the one SB 951 requires is written.

Bring Alex in to help your leadership team decide what to start counting before 2027 →

← Back to Blog
Head portrait of Alex Goryachev
Alex Goryachev

WSJ bestselling author of Fearless Innovation · Built and ran Cisco's Global Innovation Centers in 14 countries · Has advised Dell's GenAI practice and the California State University system

Work with Alex

Bring this thinking to your organization

Alex works with executive teams at global enterprises on AI strategy, governance frameworks, and organizational readiness. Available for keynotes, C-suite workshops, and advisory engagements.

Check Alex's availability

Replies within one business day.
Prefer email? [email protected]
Thank you — Alex will reply within one business day.
Oops! Something went wrong while submitting the form.