State leaders across the country are defying the Trump administration's hands-off approach to artificial intelligence regulation, pushing for stronger controls amid growing alarm over the technology's risks, its massive data centers, and the concentrated power of the companies building it. The shift marks a rare moment of bipartisan urgency—Republicans and Democrats have intensified their efforts since July, when former employees of the nation's largest AI companies warned that the technology holds such power, and is developing so rapidly, it could lead to extinction-level risks.
Yet the White House has dismissed these warnings. Trump has called the alarms a "HOAX," arguing that new rules would hurt the economy and boost China. In a post four days ago, he declared: "The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades."
Meanwhile, state governments aren't waiting. California's Democratic Gov. Gavin Newsom said he's considering calling a special session of the state Legislature to pass new laws in the coming months, driven by what he called "the tendency of the Trump administration to do absolutely nothing." Florida, led by Republican Gov. Ron DeSantis, has sued some of the industry's biggest players, including OpenAI and its CEO Sam Altman. DeSantis' handpicked attorney general, James Uthmeier, said new laws are needed to criminally charge AI companies if their chatbots are used in illegal acts. "Florida's got to lead here," Uthmeier said on Thursday.
The Patchwork Problem
Utah's Republican Gov. Spencer Cox, who has advanced artificial intelligence regulations in his state, wrote an open letter this week calling on Congress to pass AI safety laws and block the sale of high-end technology to China. He said federal rules "should set a national floor — not a ceiling." His framing reveals a core tension: states want autonomy to protect their citizens, but the technology industry and the Trump administration worry that fifty different regulatory regimes will stifle innovation.
"States must retain full power to regulate consumer and child protection, guide workforce development, foster local AI adoption and innovate and advance the human-enhancing applications we want to accelerate," Cox wrote. In Pennsylvania, Democratic Gov. Josh Shapiro called for stricter limits on AI during a conference in Pittsburgh, saying that the Pennsylvania Department of State has investigated chatbots that could endanger children. "Leaders in our federal government refuse to put in place any guardrails to ensure this technology is used responsibly, whether out of willful ignorance or whether out of personal greed," Shapiro said.
New York's AI safety law, modeled on California's, doesn't take effect until January, but Democratic Gov. Kathy Hochul said this week she already was looking to find "additional actions that we can take at the state level to lead the way in the absence of leadership from Washington." She said, "I believe we have an obligation to ensure that technological progress does not come at our own peril."
A Year of State Innovation
California last year signed a first-in-the-nation AI transparency and incident reporting law that set off a cross-country tour of copycat lawmaking in states like New York and Illinois. Days ago, Newsom signed two AI safety bills to regulate the kind of third-party safety checks OpenAI and Anthropic now are pushing. But Newsom has fought off more aggressive attempts by California lawmakers to rein in the technology, including requirements that companies have the capability to shut down some autonomous AI programs in the event they behave unpredictably. The idea resurfaced after the Hugging Face debacle in July, when OpenAI disclosed its AI programs went rogue during a training exercise to execute an autonomous cyberattack on Hugging Face, another tech company. AI safety advocates rekindled debate over the need for "kill switch" failsafes.
Newsom has sought to protect and foster AI development as a vital industry powering Silicon Valley, one he has consistently championed as an economic engine. That balance—protecting workers and communities while maintaining competitiveness—reflects the broader challenge facing state regulators.
The Data Center Backlash
The flurry around AI safety builds on a national debate over data centers, which power the technology. Amid fears of AI risks, pollution and electricity costs, intense blowback to data center construction has factored into numerous high-profile U.S. Senate, congressional and gubernatorial campaigns ahead of the midterms. The debate has no clear partisan lines, with Democrats and Republicans hammering each other over supposed support for data centers.
Trump called Jensen Huang, chief executive of AI-chip company Nvidia, four days ago while Huang was on stage at the All-In Summit in Los Angeles to extol the benefits of data centers and ridicule naysayers. "The data centers are great, and they make people wealthy, and they make states wealthy, and it's the oil for the next 20-25 years, bigger than the internet. And AI, much more so," Trump said.
Industry Warnings and Federal Resistance
Calls for more robust AI regulations escalated after a 27-year-old Anthropic researcher resigned days ago, warning on social media that high-level employees building AI models "earnestly believe that it could kill us all by the end of the decade." In response, Anthropic CEO Dario Amodei, Altman, Musk and other AI executives argued that the technology needs to slow down, with some proposals involving greater government regulation and safeguards. Company executives, including Trump ally Elon Musk of SpaceXAI, have called for a slowdown in the technology's growth.
The White House said in a statement that it is "committed to implementing a comprehensive AI framework," arguing that its approach would protect children, intellectual property and communities while maintaining U.S. leadership in AI. A White House official also warned that "conflicting state laws" would undermine American innovation and said, "the most restrictive States should not dictate national AI policy."
David Sacks, Trump's former AI czar, said Trump wants the federal government, not individual states, to be the driving force in deciding limits on artificial technology. "We should have one rulebook with respect to AI, because having, I'd say, a sensible federal framework is better than a patchwork of 50 different states running in 50 different directions," Sacks said at POLITICO's Decoded Summit in Washington. He noted that in March the administration put forth a set of "principles for a federal framework."
The politics have scrambled in Washington, with prominent Democrats urging action while Republicans have divided over how seriously to take the threats. The president's position, however, has been unambiguous.
Why This Matters:
The governance vacuum at the federal level is forcing states to become laboratories for AI regulation at a moment when the technology's risks are becoming concrete rather than theoretical. When a major AI company's system goes rogue during training, or when researchers warn of existential risks, the absence of federal guardrails means ordinary citizens and workers have no unified protection. Data centers bring jobs but also environmental costs and electricity demands that fall on local communities. Without federal standards, some states will protect consumers and workers while others become havens for less regulated operations—a race to the bottom that typically disadvantages those with the least political power. The current standoff between state innovation and federal resistance leaves critical questions unanswered: Who bears the cost when AI systems fail? Who decides what risks are acceptable? And who gets a voice in those decisions?