California’s coffers got a noticeable lift from the latest rush into artificial intelligence — but the boost is concentrated, fragile and tied to a lot of unanswered questions.
A revenue spike built on stock options
A new analysis by the Legislative Analyst’s Office found that withholding from employee stock-option income at major tech firms accounted for roughly 10% of all state income-tax withholding in 2025. That’s up from a bit over 6% just three years earlier. The LAO’s tally focuses on big players — Apple, Google, Nvidia, Broadcom and Meta — and adds other substantial withholders such as Intel, Cisco and AMD.
Why does that matter? Personal income tax is California’s single biggest revenue source. When a handful of companies see their share prices soar, a relatively small group of employees cash in stock and the state collects a big slug of withholding. When share prices sag, the reverse happens fast.
LAO principal analyst Chas Alamo told CalMatters: “We’re seeing a real boost to income-tax receipts because of this — for a relatively small number of employees. If the AI market were to deteriorate, we could see these withholdings decline.”
Jobs aren’t following the money
You might expect a boom of this scale to be accompanied by rising headcounts and higher wages. Instead, labor-market signals look mixed at best. The Bay Area saw fewer tech jobs from September 2024 to August 2025; information-industry employment slid about 1.3% over that period and professional services dropped 1.5%. By September, California’s unemployment rate ticked up to 5.6% — the highest among states.
Some large employers have explicitly tied layoffs to AI-driven restructuring. Salesforce and other firms pointed to productivity gains from automation as part of the reason they cut thousands of jobs.
That dissonance — soaring valuations without broad-based hiring — is why some economists and think tanks worry the state’s fiscal picture rests on shaky ground.
Bubble chatter, and voices on both sides
Is this a bubble or the start of a long-term transformation? The answer divides experts. Optimists such as Nvidia’s Jensen Huang and analyst Dan Ives see an extended multi-year buildout, likening today’s moment to the early mainstreaming of the web in the mid‑1990s rather than the froth of 1999–2000.
Skeptics point to classic bubble signals: frenzied investor demand, sky-high private valuations and spending that sometimes outpaces demonstrable earnings. PitchBook and Allianz Trade have flagged crowded niches — like AI scribes for healthcare or personal-assistant bots — where differentiation and sustainable business models are thin. There are also practical worries about energy: AI’s hunger for compute has spurred data center growth that strains local grids and pushes utilities planning into the spotlight.
Those infrastructure pressures help explain why companies are experimenting beyond the ground floor — see projects like Google’s Project Suncatcher that imagine placing AI capacity in atypical locations.
Meanwhile, the broader scientific and policy conversation about whether AI is truly approaching human-level capability continues. That debate, with its mix of hype and sober caution, matters because it shapes investor behavior, corporate strategy and the public policy response — all of which feed into how sustainable today’s revenues might be. For deeper reading on that debate, see AI’s Tipping Point.
Why Sacramento is paying attention
California faces a projected near-$18 billion budget shortfall this year. That gap makes the state particularly sensitive to shifts in the tech sector’s fortunes. Economists warn that revenue swings from a concentrated source can force difficult trade-offs: cuts to services, deferred investments, or politically painful tax changes.
There’s also a political angle. California leads the states pushing its own AI regulations and is preparing to push back against federal efforts that would preempt state rules. How regulators handle AI could influence how fast companies expand, where they locate compute and hiring, and therefore how reliably the state can count on that income stream.
Supporters of a light-touch approach point out that tech revenue underwrites much of California’s social safety net and public services. As Kaitlyn Harger of the Chamber of Progress put it, a healthy tech sector funds schools, health services and state payrolls. Opponents worry that leaning into a single industry leaves the state exposed to cycles it cannot control.
No neat ending emerges from the numbers. California is enjoying a boom for now — its budget lines brightened by stock-option windfalls — but the underlying labor market, infrastructure limits and an unsettled debate about how real and durable the AI revolution is leave a lot of uncertainty. Policymakers in Sacramento know the math: a hurry of equity gains can fill a gap today and empty it tomorrow.
If the last few years taught anything, it’s that when revenue arrives in a rush, planning for the possibility that it could ebb should be part of the work. The question for California is how long it wants to build big bets on an industry still sorting out what “mature” looks like.