California has become the front line in a looming fight over artificial intelligence regulation, and most of the world’s largest AI companies sit right in the blast radius.
Lawmakers in the state aren’t backing off, even after President Donald Trump threatened to punish states that move ahead with their own AI rules.
Democratic legislators argue that unrestrained AI already harms mental health, especially among children, and that waiting carries its own risks.
Trump’s December executive order, which aims to withhold federal funding from states that regulate AI, hasn’t changed that view. Companies have ramped up lobbying. Legislators say the pressure is loud, not decisive.
Assemblymember Rebecca Bauer-Kahan said industry spending hasn’t altered her position. She plans to reintroduce a bill barring minors from using so-called companion chatbots that simulate human relationships, a proposal vetoed last year by Governor Gavin Newsom. She said when harms become visible, lawmakers answer to voters, not lobbyists.
California isn’t alone. Legislators in both red and blue states have begun pushing back against Trump’s stance as reports grow of users forming unhealthy or delusional attachments to chatbots.
California’s influence still carries weight. In September, Newsom signed a law requiring AI developers to disclose safety protocols, and lawmakers in New York are already using that framework while drafting their own rules.
The regulatory push collides with a blunt economic reality. AI has turned into one of the state’s strongest revenue engines inside a roughly $320 bn budget. The Silicon Valley firms under scrutiny hold a combined market value above $15 tn.
Income tax withholdings tied to AI-driven growth at companies like Apple, Nvidia, and Alphabet now deliver an estimated $10 bn a year to California’s treasury.
The state’s Legislative Analyst’s Office has called it the lone bright spot in an otherwise bleak fiscal outlook.
Newsom, a Democrat with national ambitions and parents watching closely, sits in the middle. Industry groups warn companies could relocate if California imposes rules they see as hostile.
They point to recent exits by Hewlett Packard Enterprise, Oracle, and Tesla.
Catherine Bracy, CEO of TechEquity, said the threat of flight is the industry’s favorite bargaining chip. And it gets played often.
Behind the scenes, negotiations have intensified. Lawmakers, lobbyists, and founders are sparring over minors’ access to AI tools and over the use of copyrighted material in training models. Voters may enter the picture next.
Common Sense Media is collecting signatures for a ballot initiative to limit underage chatbot use. OpenAI plans to put a competing measure on the November ballot. Each needs more than 500,000 valid signatures by June.
Common Sense founder Jim Steyer said he has held talks with OpenAI executives to find common ground. He calls California the de facto center of major tech regulation in the US. Steyer argues public opinion sits with child safeguards, and that momentum favors limits, not delay.
OpenAI said it is exploring stronger teen safety protections, including age prediction and added parental controls, building on existing California standards.
Tech firms are also escalating political spending. They highlight AI’s role in funding schools and social services. Adam Kovacevich, CEO of Chamber of Progress, said the industry isn’t asking for tax breaks, only recognition for its contribution to the state’s safety net. The group is backed by Andreessen Horowitz, Apple, OpenAI, and others.
Kovacevich pointed to a moment in 2024 when Nvidia’s stock surge flipped California’s budget outlook from deficit to surplus. He questioned why lawmakers would risk harming what he called the golden goose.
Chamber of Progress opposes Assembly Bill 412, which would force companies to disclose copyrighted material used to train generative AI models.
The group estimates the bill could cost California at least $381 mn in lost revenue. Hollywood unions back the proposal, arguing creators need enforceable protections.
Elsewhere, Meta Platforms has seeded a super PAC focused on California’s AI sector with $20 mn. The California Chamber of Commerce warns that broad bans on minors’ access to chatbots could push AI companies out of the state altogether.
Lawmakers return to Sacramento on Jan. 5 and insist safety rules won’t choke innovation. One bill, SB 300, is set for amendments to block sexually explicit chatbot content for minors.
Another proposal expected this month would impose a four-year moratorium on selling AI-powered chatbot toys to children. Whether the industry blinks first, or the state does, remains an unfinished thought.
