California is taking an unusually direct shot at one of AI’s ugliest side effects: job displacement. Governor Gavin Newsom has signed what his office says is the first state-level executive order in the United States aimed at cushioning workers if automation starts replacing them, and the move mixes retraining, subsidies, and a fresh look at how society shares the gains from AI.
The California AI worker plan sets up a cross-agency effort involving state officials, researchers, labor groups, and AI companies. That sounds broad because it is broad, which is probably the point: the state is trying to get ahead of a labor problem that existing unemployment systems were never designed to absorb at scale.
California AI worker plan: subsidies, retraining, and ”universal capital”
The initiative includes financial incentives for employers that keep people on the payroll instead of swapping them for AI tools, plus expanded retraining programs aimed at office jobs most exposed to automation. California is also revisiting the idea of ”universal capital,” a concept that would give citizens stakes in investment funds or other assets rather than relying only on traditional welfare checks.
That last piece is where the politics get interesting. In Europe and parts of the U.S., governments have toyed with versions of universal basic income for years; California is pushing a more asset-focused variation that tries to turn workers into partial owners of the economy that is reshaping their jobs.
The warning sign behind the order
Newsom says the normal unemployment system won’t be enough if AI starts chewing through white-collar work faster than people can retrain. His pitch draws on bleak forecasts from AI executives and critics alike, including the view that a huge share of office roles could disappear within a few years if companies keep automating at the current pace.
He also took aim at tax rules that, in his view, reward automation while making human labor look expensive by comparison. That complaint is not new, but it is becoming harder for policymakers to ignore as firms use AI to trim costs first and ask social systems to deal with the fallout later.
A state experiment with national stakes
For now, California is acting alone, which may be the most revealing part of the story. The state has the political weight and tech concentration to try a labor policy test run, but without a federal framework the real question is whether any of this can scale beyond one very wealthy state.
If the plan gains traction, expect other states to borrow the language before they borrow the money. If it flops, California will still have done something useful: it will have put a price tag on a question most governments are still trying to dodge.