California flipped on a public dashboard this week that tries to answer a question everyone argues about and almost nobody has data for: is AI actually taking jobs? Governor Gavin Newsom's office announced the California AI-Unemployment Tracker on June 25, built with the California Policy Lab at UCLA and the state's Employment Development Department. It's the first state tool of its kind, the administration says.
How it actually works
The method is the interesting part. It takes the state's monthly unemployment-insurance claims and tags each one by how exposed the worker's old job was to AI. Customer service reps and software developers score high. Heavy-goods drivers and nursing assistants sit near the bottom. Workers self-report their occupation from a standard list when they file, which the researchers then map to an exposure score.
That self-reporting is also the catch. Nobody verifies the job titles. The data misses gig workers, the self-employed, and anyone who never files a claim at all. The lab is unusually upfront about this. The tracker is a signal, not proof, and it can't say AI caused any single layoff. No dataset can.
What the first numbers say
Not much of a jobpocalypse, so far. Newsom's office says there's no evidence of rising statewide unemployment claims in AI-exposed occupations through May 2026.
"Right now, we are not seeing evidence of large-scale AI-related layoffs in California's labor market." That's co-author Dr. Ben Hyman, and the data mostly backs him, though he immediately points to the exceptions.
The exceptions are where it gets pointed. Claims from college-educated workers in high-exposure jobs climbed after ChatGPT-3.5 came out in 2022 and stayed elevated. The San Francisco Bay Area saw a sharp, sustained increase among high-exposure workers over the same stretch. So the headline number is calm while the regional detail is not, which is roughly what you'd expect if early disruption hit the tech-heavy parts of the state first.
One thing worth chewing on: the researchers strip out the pandemic years because that surge would swamp everything else. Reasonable. It also means the trend you're looking at is built on a deliberately edited timeline.
The politics aren't subtle
Newsom is widely expected to run for president in 2028, and "we're not watching from the sidelines" is the kind of line that plays well when voters are nervous about AI. He's not alone in the scramble. Missouri's Josh Hawley introduced bipartisan legislation requiring companies to report AI-related layoffs. Bernie Sanders keeps raising the alarm. A New York lawmaker floated an "AI Dividend" tied to displacement.
California has a real reason to go first, beyond optics. It hosts most of the firms building advanced AI and carries one of the highest unemployment rates of any state. Center of the boom, natural test case for the fallout.
The dashboard updates monthly, so the next read lands in roughly four weeks. If the Bay Area signal keeps climbing, EDD says it'll route retraining and job-search support toward the occupations showing strain.
