My friend Ben Hyman, my coauthor Jacob Morris, and their team at the California Policy Lab — in partnership with the state’s Employment Development Department — just dropped a very valuable tool: the California AI-Unemployment Tracker (CAIT).
Up until now, as a state or a society, we had no government data actively tracking the relationship between AI exposure and unemployment insurance claims. Now we do.
And because of it, we know — with unemployment claims data from as recently as May 2026 — that high-exposure roles are experiencing an increase in unemployment claims. In the Bay Area, the observed AI exposure score has been increasing more than it has in the rest of California. The same is true of Professional Services compared to other occupations.
So here’s the question I keep turning over:
Is the Bay a harbinger? Or is Professional Services simply doing what we expect other occupations to do?
I don’t know. But I do think this is exactly the kind of signal that supports ongoing monitoring, and I’ll be keeping an eye on it — alongside the Stanford Digital Economy Lab’s AI Economic Indicators.
What I want to emphasize is how big a deal it is that this exists at all. This is a huge endeavor by the state of California: matching initial unemployment claims back to 2017 against occupational AI exposure, and putting it out in the open for anyone to scrutinize. That’s the infrastructure we need if we’re going to understand what AI is actually doing to the labor market — not anecdotes, not vibes, but data we can return to month after month.
Celebrating great tools made by great people that help us understand the labor market. Go explore the tracker.