States Are Embracing AI Regulation, Defying Tech Industry and White House

Has Big Tech finally overplayed its hand? Having lost two separate and deeply significant cases about liability for content in federal court last week, the industry is increasingly being confronted by states - even in conservative southern and western regions - over the threat posed by AI. 

Between data centers, electricity costs, child safety, intellectual property, privacy, as well as the loss of millions of jobs which the industry has unwisely touted as one of AI's premier benefits, the cascade of concerns about AI by otherwise tech indifferent and apolitical Americans is skyrocketing, to the point where state legislatures are passing dozens of bills regulating the technology. And as if the perceived threats to individuals' safety and livelihood weren't enough, the high-handed disdain of billionaire tech bros to those concerns has further inflamed the issue. That the industry has obsequiously fallen all over itself to kowtow to the Trump administration has added to the growing negative public reaction (Amazon's $75 million 'investment' in an adoring movie about Trump's wife, Melania, being merely the most craven) and to a sense that this White House is favoring the industry over the voters. Twenty-five years after dotcom, with every company now being dependent on tech and almost every individual wielding a smartphone, the populace understands both the good and the bad. They are no longer willing to take the industry's word for anything, believing, based on hard-won experience, that it is almost entirely self-serving. The result is increasing demand for regulation - and the states are responding. JL

Cecelia Kang reports in the New York Times:

The battle over who should regulate A.I. is turning into an epic clash, as anxiety soars over its effects on jobs, education, national security and child safety. States have increasingly taken matters into their own hands, introducing dozens of bills this year to put guardrails around A.I. More than 100 state laws ban chatbots for young users, mandate system testing for security risks and protect copyrighted materials from scraping into A.I. systems. The fight has taken on new vehemence as power-guzzling data centers and rising electricity costs, as well as A.I.’s disrupting the broader economy, have taken center stage in elections. OpenAI, Google and Meta have thrown hundreds of millions of dollars for candidates favorable to the industry. Trump has made it clear US companies should have free rein in A.I., and that state rules hold them back. "They don’t want Big Tech inconvenienced.”

This month, President Trump warned states not to get involved in regulating artificial intelligence. In a set of policy guidelines, the White House said a “patchwork of conflicting state laws would undermine American innovation and our ability to lead in the global A.I. race.”

But on Monday, Gov. Gavin Newsom of California, a Democrat, defied Mr. Trump by issuing an executive order that requires safety and privacy guardrails for A.I. companies contracting with the state. He also said he would fight to preserve California’s laws that provide safeguards against A.I.-related catastrophic harms, scams and risks for children.

“We’re not going to sit back and let that happen,” Mr. Newsom said in a statement, referring to Mr. Trump’s effort to supersede state action. “It’s taking a sledgehammer to the very protections Californians rely on.” 

The battle over who should regulate A.I. is turning into an epic clash between Mr. Trump and the states, as anxiety has soared over the technology’s potential effects on jobs, education, national security and child safety. States have increasingly taken matters into their own hands, introducing dozens of bills this year to put guardrails around A.I. Already, more than 100 state laws ban chatbots for young users, mandate system testing for security risks and broadly protect copyrighted materials from being scraped into A.I. systems.

But Mr. Trump has made it clear that American companies should have mostly free rein in the global race to dominate A.I., arguing that state rules will hold them back. The White House’s policy guidelines for federal legislation this month called for blocking state laws on A.I., while recommending some safeguards for children and consumer protections for energy costs.

Mr. Newsom’s executive order was just the latest skirmish. In February, the White House issued a warning letter to Utah legislators that derailed an A.I. transparency and child safety bill. This month, Gov. Ron DeSantis, a Republican, failed to get an A.I. bill passed in Florida after Mr. Trump’s opposition to state intervention. In August, Colorado delayed an A.I. law after intense criticism from the White House and tech companies.

“A.I. is like the Super Bowl of public policy because it is hyperlocal, national and international in significance,” said Kevin Frazier, an adjunct research fellow at the Cato Institute, a libertarian think tank. “And everyone feels like they have the right to do something about it." The state and federal war follows years of tension during which a deeply divided Congress has failed to pass any laws regulating Big Tech. State legislators say they have been forced to take the lead in protecting children and strengthening privacy for consumers of social media.

“There isn’t a single serious policy person who thinks what the White House is doing will protect consumers,” said Alex Bores, a Democrat in the New York State Assembly who co-sponsored an A.I. law that requires companies to report safety incidents. The fight has taken on new significance as topics like power-guzzling data centers and rising electricity costs, as well as concerns about A.I.’s ability to disrupt the broader economy, have taken center stage in some state and local elections. OpenAI, Google and Meta have thrown hundreds of millions of dollars into lobbying for candidates favorable to the industry.

The debate has also caused tension within the Republican Party, which includes billionaire tech leaders as well as social conservatives and economic populists who worry about the technology’s harms. “There is not even a real attempt to legislate on the federal level, and it’s as if they don’t want Big Tech to be inconvenienced,” said Louis Blessing, a Republican state senator in Ohio. “I find that deeply offensive.” Mr. Blessing recently wrote a letter to the White House, signed by dozens of Republican state lawmakers from across the country, calling for the White House to let states regulate A.I.

A White House official said in a statement that its proposed framework for legislation pre-empting state laws would include protections for children, force companies to pay for energy costs associated with data centers and consider copyright protections.

David Sacks, Mr. Trump’s top adviser on A.I., has said the legislative framework has some compromises to protect consumers. “I think there is a very good chance Congress will act on this framework and we’ll get meaningful legislation in the next few months,” Mr. Sacks said in an interview on Bloomberg Television.

States started their push to regulate A.I. soon after OpenAI introduced ChatGPT to the public in November 2022. In 2024, California and Colorado were among the first to introduce bills that require testing and transparency of A.I. security and safety. Nearly 40 states introduced A.I. bills curbing deepfakes and bolstering child safety and testing for national security last year.

President Joseph R. Biden Jr. signed an executive order in October 2023 requiring safety standards for government use of the technology, but after Mr. Trump returned to office last year, he signed an executive order overturning it. In an interview this month, Mr. Sacks warned about the rising threat of China and said regulations would impede American technology companies. The battle over state regulations erupted soon after Mr. Trump’s executive order. In May, Speaker Mike Johnson, Republican of Louisiana, helped pass a House proposal for a decade-long moratorium on state A.I. laws.

The venture capital firm Andreessen Horowitz and A.I. companies had aggressively lobbied for the bill. They next turned to Senator Ted Cruz, Republican of Texas and chair of the Commerce, Science and Transportation Committee, to push forward a companion bill in the Senate. He tried to add the same proposal — deeply unpopular to many senators — to a budget-reconciliation bill.

Gov. Sarah Huckabee Sanders of Arkansas, a Republican, wrote in The Washington Post, “Americans are at risk from bad actors in the A.I. industry until lawmakers are allowed to establish basic rules and fairness — and 10 years is too long to wait.”

The bill was defeated in a 99-to-1 vote in July.

States continued to push forward with their own legislation. In September, Mr. Newsom signed a new bill requiring transparency and safety reporting from the biggest A.I. companies.

Mr. Trump continued to push back. In December, he signed an executive order, written by Mr. Sacks, that ordered states to stop passing A.I. laws. It said the Justice Department would sue them and pull federal funding if they did. States ignored the warning. In January, Illinois amended a human rights law to include requiring disclosures when A.I. is being used in job recruitment. In Minnesota, State Senator Erin Maye Quade, a Democrat, has proposed seven A.I. bills this year, including a child safety bill related to chatbot use.

In Utah, State Representative Doug Fiefia, a Republican, introduced a bill in January, known as H.B. 286, that would require A.I. companies to publish safety plans and child safety measures. After a key committee advanced it, the White House weighed in. “We are categorically opposed to Utah H.B. 286 and view it as an unfixable bill that goes against the administration’s A.I. agenda,” the White House Office of Intergovernmental Affairs wrote to the Utah Senate’s majority leader. The full Legislature did not pass the bill.

“It was concerning to see unelected bureaucrats trying to discourage states from addressing issues that directly affect our families and communities,” Mr. Fiefia said. He added that he would reintroduce his bill in the state’s next session. After the White House released its blueprint, some Republicans have promised to pass federal legislation to carry out Mr. Trump’s agenda. But Democrats promised to fight it.

“Congress must not only reject Trump’s offensive proposal but also continue blocking Republicans’ bad-faith effort to undermine states,” Senator Edward J. Markey, Democrat of Massachusetts, said.

Mr. Newsom’s executive order on Monday will force A.I. companies contracting with the state to report policies to protect public safety. They will also have to report how they prevent surveillance and the distribution of illegal content, like child sexual abuse material.

“We’re going to use every tool we have to ensure companies protect people’s rights, not exploit them or put them in harm’s way,” Mr. Newsom said.