As voter anxiety over AI grows, candidates stake out positions

Ah, the 2026 election cycle, and every candidate has suddenly discovered artificial intelligence — an issue which, until recently, they were happy to leave to the people who make it and the people who ask it to compose wedding speeches. But voter anxiety over AI grows, so the platform is being assembled out of whatever polls well: ban AI from replacing teachers, tax AI services and products, regulate large frontier models. That's the whole plan. It is like fighting a meteor with a rain poncho and a politely worded letter, except the letter has been focus-grouped.

Taxing AI services is a beautiful, stupid idea — the logic, presumably, being that if the machines take all the jobs, we can at least make them pay VAT. The teachers' ban is more endearing, a candle in a hurricane, except it protects one lucky profession while the rest of the workforce gets to find out just how enthusiastically their jobs can be automated. And 'regulating large frontier models' sounds admirably serious until someone asks what 'large' means, at which point you get the kind of answer a fortune teller gives: confident, vague, and different every time. Governor races including California — the state that actually grows the labs — will now perform this ritual in public.

As for the anxiety that started it all: the candidates have staked out their positions, and nothing reassures a terrified electorate quite like a tax that will be litigated for a decade and a regulatory framework that will be outdated before the ink dries. The gap between what AI is doing and what any government can do about it remains roughly the size of California, which is apt, because that's where the conversation is happening. Teachers can sleep soundly tonight — right up until a school district discovers that a chatbot costs less than a pension and runs 24 hours a day. Same time tomorrow, folks.