Someone needs to stand up....

State attorneys general seize on AI as Washington stalls

Drawing on their experience taking tech companies to court, red and blue states’ top lawyers are campaigning on a more aggressive stance against AI companies.

By Casey He

State attorney general elections are becoming the latest flashpoint in the fight over artificial intelligence regulation, as it moves from a gridlocked Washington to the campaign trail.

As distrust in the AI industry and its ability to self-regulate grows — with another jolt from dire warnings raised by industry leaders and insiders last month over AI’s existential risks — it’s presenting an especially ripe opportunity for the states’ top lawyers.

In red and blue states, attorney general incumbents and candidates are vowing to take on AI and the powerful companies behind it. Their pitch: They have the roadmap because they’ve already taken Big Tech to court and won, including a landmark $17 billion settlement with Meta in August.

“I think we’re just seeing so many instances of AI, Big Tech companies that don’t seem to think they play by the same rules as everybody else,” said Arizona Attorney General Kris Mayes, a Democrat up for reelection in the perennial battleground. “I believe that Americans feel that and want to see us stand up for them and bring some of these companies to heel.”

For many of the attorneys general, the legal battles against Big Tech have been both a cautionary tale of too little regulation, too late, and an argument for states to retain the power to regulate powerful companies.

“If Congress’s slow reaction for social media is similar to AI, then we kind of know that this will ultimately end up at the state level. So why not get a head start?” said Katherine Robertson, the longtime chief counsel to Alabama’s Republican Attorney General Steve Marshall who is now the favorite to succeed him in November. “The last thing we want to do is wait five years to then try to react to something that Congress hasn’t handled.”

Robertson, who’s now leading Alabama’s investigation into OpenAI, said she expects the office to be more aggressive toward AI companies to make sure their products are safe, especially for young people, if she’s elected.

The attorneys general have been at the forefront of the federalism tug-of-war with Washington over AI regulation since last year. After Congress took up the idea of a federal moratorium on state and local AI laws, 40 bipartisan attorneys general signed a letter to congressional leaders opposing the effort.

President Donald Trump seems especially adamant about stopping states from making their own rules. In a December 2025 executive order, the president directed the federal government to withhold funding and possibly litigate against states with “onerous and excessive” AI regulations. The Trump administration has maintained that a state-led patchwork of laws could undermine the industry and threaten national security.

“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,” David Sacks, the president’s former AI czar, said at POLITICO’s Decoded Summit in Washington last month.

Yet for the attorneys general, the White House’s call, paired with congressional inaction, increasingly rings hollow.

South Carolina Attorney General Alan Wilson, who’s now the Republican nominee for governor, was among the bipartisan attorneys general opposing the congressional preemption. He has focused particularly on AI-generated child sexual abuse material, having backed state and federal efforts to crack down on the issue.

“There is absolutely an important role for the federal government, especially when it comes to national security, interstate commerce, American competitiveness, and threats that cross state lines,” Wilson’s campaign said in a statement. “But federal standards should not prevent states from protecting their own citizens. The Attorney General has believed from the beginning that getting AI policy right will require both federal and state action. That position has not changed.”

In New Mexico, Attorney General Raúl Torrez, who is shepherding two major cases against Meta involving child safety and data privacy, said his experience has made him lose confidence that Big Tech, with the federal government’s encouragement, can be trusted to regulate itself.

New Mexico chose to take Meta to trial in the child-safety case and won a nearly $1 billion judgment earlier this year. Meta has said it would appeal the ruling.

“I just don’t think that anybody has any faith in leaving an industry to drive technology that is this consequential without real government accountability and oversight behind it,” Torrez, a Democrat who’s running for reelection this year, said.

Torrez has also been using his perch to push the state legislature to pass AI regulation, including a new proposal he unveiled last week targeting frontier AI models.

States have enacted more than 100 AI laws this year, as their legislatures rush to fill the regulation void. Even some of the most vocal proponents of federal preemption acknowledge that the dearth of federal guidelines has fueled the surge in state action.

“They make a fair point that there’s been limited action at the federal level because Congress is unfortunately a dysfunctional mess that seemingly has become a bit of a non-actor on technology policy,” said Adam Thierer, a resident senior fellow for technology and innovation at the libertarian R Street Institute. “In the meantime, they’re going to rush forward with this because it’s the politically opportunistic thing to do.”

The attorneys general’s focus on AI in campaigning comes as many are already litigating against the technology.

In California, Attorney General Rob Bonta has probed leading AI companies. In New York, Attorney General Letitia James hasdirected AI safety whistleblowers to her office’s tip line. In Florida, Attorney General James Uthmeier has sought to stop OpenAI from advancing its models in a lawsuit against the company. In Texas, Attorney General Ken Paxton has investigated several AI firms.

The AI industry has been among the biggest political players in this year’s federal elections, with its rival super PACs pouring in tens of millions into congressional primaries. But state-level races have attracted comparatively less spending.

The Anthropic-backed Public First Action has contributed to the Statehouse Fund, which spent about $500,000 supporting Delaware Attorney General Kathy Jennings in her primary. A spokesperson for Public First Action said it has not directly spent on any attorney general races, but that the group welcomed the attorneys general’s scrutiny.

“The fact that candidates in both parties are championing AI accountability on the campaign trail tells you where the public is,” the group said in a statement. “Right now, AGs are the officials who can actually bring a case. We’d like to see Washington catch up, but until it does, the state AGs are the front line.”

Leading the Future, a rival super PAC backed by OpenAI executives and others, did not respond to a request for comment, nor did The Statehouse Fund.

A July poll by the center-left think tank Third Way found 51% of voters say the government is best suited to regulate AI and other tech sectors, though there was no consensus on whether Washington or state and local governments should lead. Just 16% say the technology should be regulated by the industry, or not at all.

“Voters want someone to do something to protect them,” said Ruth Whittaker, Third Way’s director of technology policy. “They don’t really care about interstate commerce and which agency has the correct authority and the correct funding to be doing it.”

And attorney general candidates say they’re ready to step up.

“This idea of preempting anybody from doing anything while the lords of industry march forward under their own auspices is ridiculous,” said Texas state Sen. Nathan Johnson, who is running to be the state’s first Democratic attorney general in decades. “It’s irresponsible, and it’s inconsistent with the obligations of government.”