By Eliyahu Kamisher, Bloomberg
This year, many of the world’s most powerful artificial intelligence companies face a pitched battle over government regulation on their home turf — California.
And even President Donald Trump’s threat to punish states that regulate AI may not stop the fight.
California lawmakers, dominated by Democrats, are determined to place guardrails on the homegrown industry, saying unfettered AI poses a mental health risk to children and adults alike. They scoff at Trump’s executive order in December to withhold federal funding from states that adopt AI rules, saying it’s their responsibility to act. And while companies have launched a lobbying blitz to block what they consider onerous regulations, legislators say they won’t be deterred.
“They’ve spent a lot of money, and their influence is real,” said Assemblymember Rebecca Bauer-Kahan. “When the harms become so salient, my colleagues are there to serve the people, and that noise doesn’t stop us.” She plans to reintroduce in January a bill that would bar minors from using “companion” chatbots that form human-like relationships, legislation that Governor Gavin Newsom vetoed last year.
Legislators in both blue and red states are bucking Trump on AI, as reports mount of children and adults developing unhealthy or even delusional attachments to chatbots. And California has a long history of passing legislation designed to serve as a blueprint for other states. In September, for example, Newsom signed into a law a bill requiring AI developers to disclose safety protocols, and New York lawmakers are now using it to develop their own regulations.
But the push for AI rules in California also faces a stark economic reality. The industry has become a growing source of tax revenue for the state’s roughly $320 billion budget.
The Silicon Valley giants targeted by lawmakers carry a combined market value exceeding $15 trillion. AI-fueled income tax withholdings from companies such as Apple Inc., Nvidia Corp. and Alphabet Inc. now deliver an estimated $10 billion to the state treasury — the “lone bright spot” in an otherwise grim fiscal forecast, according to California’s nonpartisan Legislative Analyst’s Office.
“Where do you think this money comes from?” Newsom said, in an interview with Bloomberg Businessweek.
That puts Newsom, a Democrat, in a difficult spot. He’s weighing a run for the White House in 2028, and anxious parents are a potent voting bloc. But industry representatives warn AI companies could flee the state if it imposes onerous rules, joining a recent corporate exodus from California that includes Hewlett Packard Enterprise Co., Oracle Corp. and Tesla Inc.
“It’s the ace up their sleeves,” said Catherine Bracy, chief executive officer of TechEquity, a nonprofit that advocates for AI safeguards. “And they play that card a lot.”
As a result, legislators, lobbyists and AI entrepreneurs are engaged in intense negotiations over children’s access to the technology as well as the use of copyrighted material in AI training. California voters may get to weigh in as well. Child advocacy group Common Sense Media is gathering signatures on a proposed state ballot initiative to limit underage chatbot use, while OpenAI wants to place its own, competing measure on the November ballot. Each side must gather more than a half million valid signatures by June to qualify.
Common Sense’s founder, Jim Steyer, has been talking with OpenAI executives, seeking a compromise. Steyer, whose billionaire brother Tom is running for governor to replace the termed-out Newsom, calls California the “de facto center of important regulation of Big Tech.”
“We appeal to their better instincts as parents, as citizens and as good people,” Jim Steyer said. “At the end of the day, we believe that California will set the standard for the rest of the nation. And we think that we actually have the momentum. We think the public agrees with us.”
OpenAI, in a statement, said the company is “exploring additional ways to strengthen teens’ safety protections, including age prediction and more parental controls, building on California’s existing standards.”Tech companies, aware of the stakes, are escalating their political engagement — and touting AI’s contributions to California’s public schools and social services.
“The AI industry is not asking for tax relief or special tax handouts in Sacramento — it just wants great appreciation for the role it’s playing in powering the state’s social safety net,” said Adam Kovacevich, CEO of Chamber of Progress, a group trying to encourage pro-tech policies within the Democratic Party. It’s backed by Andreessen Horowitz, Apple, OpenAI and others.
“There was a moment [in 2024] where the state’s budget outlook swung from deficit to surplus strictly on the basis of unexpected revenue from Nvidia’s stock performance,” Kovacevich said. “I don’t understand why the legislature would do anything to harm the golden goose.”
His group is pushing back against a bill, AB 412, that would require companies to disclose which copyrighted material they used to train a generative AI model. According to their analysis, the legislation would cost the state at least $381 million in lost revenue. The bill is backed by Hollywood unions seeking to ensure protections for artists and writers.
Meta Platforms Inc., meanwhile, seeded a super PAC focused on the state’s AI sector with $20 million. And the California Chamber of Commerce, which represents many major tech firms, is warning that overly restrictive rules banning minors’ access to chatbots could “easily push the AI industry out of California”
Legislators, who reconvene in Sacramento on Jan. 5, insist safety measures won’t stifle AI innovation. They see their role as setting sensible ground rules for a fast-growing, globally important industry. For example, one bill set to be amended in January — SB 300 — seeks to ensure chatbots do not provide sexually explicit content to minors. Another proposal that will be introduced this month would place a four-year moratorium on the sale to minors of toys powered by AI chatbots.
“I remember the auto industry talking about how airbags in cars and seatbelts were going to be the end of the industry, and they were going to stifle productivity and innovation,” said Senator Steve Padilla, who wrote legislation signed into law by Newsom last year requiring companies to inform users that they are interacting with artificial intelligence and report suicidal behavior. “I think most people understand that there is a responsible, reasonable place for appropriate safeguards to protect everyone’s best interests.”
–With assistance from Shirin Ghaffary.
More stories like this are available on bloomberg.com
©2026 Bloomberg L.P.
Leave a Reply