Is AI Taking Jobs? California Built a Tracker to Find Out — Here’s What It Shows

For three years, the question has run on vibes. Someone loses a copywriting gig, a forum fills with worry, a CEO boasts about replacing a team with software, and the headlines swing between “AI will take half of all jobs” and “AI changes nothing.” Almost none of it rested on data that tracked actual workers losing actual jobs.

California decided to measure it. On June 25, 2026, Governor Gavin Newsom’s office launched the California AI-Unemployment Tracker, the first state system built to watch for AI-related job loss as it happens. The early numbers complicate both the panic and the dismissal. Here is what the tracker found, how it works, and what it can and cannot tell you.

What the Data Shows so Far

The headline finding cuts against the loudest fears. Through May 2026, the statewide data and unemployment claims show no surge among workers in jobs most exposed to AI, according to the California Policy Lab, the UCLA-based research center that built the tracker with the state’s Employment Development Department. The share of claims coming from AI-exposed workers has not risen by a statistically meaningful amount compared with the years before the pandemic, even though ChatGPT reached the public back in late 2022.

Also ReadEntertainmentAlan Ritchson to Star in Detroit Crime Thriller Motor City→

The reassuring read stops there. The same data flags signals worth watching. Ben Hyman, a senior researcher at the California Policy Lab and a co-author of the report, put it carefully: the state is not seeing large-scale AI layoffs, but it does see patterns in certain regions like the Bay Area, in tech-heavy sectors, and among highly AI-exposed workers who hold college degrees.

Three groups stand out in the early numbers. College-educated workers in high-exposure occupations have filed relatively more claims. Workers between 25 and 35 show up among the more exposed groups. And the San Francisco Bay Area, dense with technology firms, records more activity than the rest of the state. The picture lands between reassuring and cautionary, which is a harder story to tell than either extreme.

How the Tracker Actually Works

Most attempts to measure AI’s effect on jobs ask employers to report it, which means the data only captures what companies choose to admit. California took a different route.

Also ReadBusinessWall Street’s Next Billion-Dollar Shift Is Already Happening→

The tracker pairs two datasets. The first is unemployment insurance claims, which the state already collects when someone files for benefits. The second is an “AI exposure” score that estimates how much a given occupation could be automated. By matching a claimant’s most recent job to that exposure score, the researchers can watch whether claims rise faster among jobs a machine could plausibly do.

The method has a built-in limit, and the researchers state it plainly. The tracker spots statistical patterns. It does not prove that AI caused any particular layoff. A spike in claims among software workers could reflect AI, or a broader tech downturn, or a hiring correction after the pandemic boom, or all three at once. Till von Wachter, the UCLA economist who co-authored the report, framed the tool as a way to replace speculation with evidence rather than to deliver a verdict.

That distinction matters when you read the next breathless headline. “Unemployment claims rose among AI-exposed workers” is not the same sentence as “AI took those jobs,” and the people who built the tracker are the first to say so.

Why California Built it and what comes next

The tracker did not appear on its own. Newsom signed an executive order in May 2026 directing state agencies to prepare workers, small businesses, and communities for economic disruption from AI, and the dashboard is one piece of that order. Stewart Knox, who leads the state’s Labor and Workforce Development Agency, described the goal as grounding decisions in data so the state can direct job-search help, retraining, and other support where claims start climbing.

The dashboard updates monthly and stays public, so workers and researchers can watch the trend lines themselves rather than wait for a press release. The underlying data is available to download.

California’s move also exposes a gap. No other US state runs a comparable system, which means most of the country is guessing. One data point from outside the US hints at how different the picture can look elsewhere: Morgan Stanley reported in February 2026 that British firms claimed an 8% net job loss from AI, roughly twice the global average that the survey found. Survey self-reporting and unemployment records measure different things, so the two numbers do not contradict each other so much as describe the elephant from different angles. The honest takeaway is that without consistent measurement, policymakers fly blind.

What this means for you

If you work in an AI-exposed field, the California data offers cold comfort and a real warning at the same time. The mass-layoff wave that dominates online debate has not shown up in the statewide numbers. The localized pressure on college-educated workers in tech-heavy areas has.

A few grounded conclusions hold up:

• One state’s claims data is not a national forecast. California’s tech concentration makes it an early indicator, not a stand-in for the whole country. Treat it as one signal among several.

• “Exposed” does not mean “replaced.” An exposure score measures what a tool could automate, not what employers have chosen to do. Many exposed jobs change rather than disappear, with AI handling parts of the work while the role survives.

• The trend matters more than the snapshot. A single month showing no surge is weaker evidence than six months of monthly updates pointing in the same direction. The value of the tracker grows as the data accumulates.

For now, the most defensible statement is the one the researchers themselves make. California sees no broad AI layoff wave yet, sees specific groups under more pressure, and plans to keep watching. Anyone claiming more certainty than that, in either direction, is selling something the data does not support.

 Sources & References

• Office of Governor Gavin Newsom — “California becomes the first state to launch a tool to monitor and track AI’s impacts on the workforce”

• California Policy Lab (UCLA) — “California AI-Unemployment Tracker (CAIT)” and associated report

• CBS News Sacramento — “California launches tool to track whether AI is costing workers jobs”

• Fast Company — “California launches a tracker to monitor AI-related job loss”

• Governing — “California Creates the Nation’s First AI Job-Loss Tracker”

• Engadget — “California launches tracker for AI-related job losses”

• Crypto Briefing — “California launches first-in-nation AI job loss tracker”

• iTech Post — “California AI Job Loss Tracker Launches to Monitor Layoffs”