Regulatory tug-of-war

Editor's note: The following text is a transcript of a podcast story. To listen to the story, click on the arrow beneath the headline above.

MARY REICHARD, HOST: Up next, the White House goes after state AI laws.

On Thursday, President Trump signed an executive order calling on Congress to create a national policy for artificial intelligence. He also commissioned the Department of Justice to create an AI Litigation Task Force…to sue states with “cumbersome regulations” on AI.

WORLD’s Mary Muncy has the story.

MARY MUNCY: President Trump says the U.S. is winning the AI race with China, but red tape at home could trip companies up.

PRESIDENT TRUMP: We have the big investment coming. But if they had to get 50 different approvals from 50 different states, you could forget it, because it's not possible to do, especially if you have some hostile—all you need is one hostile actor, and you wouldn't be able to do it.

Since 2024, states across the country have introduced more than 1,200 AI bills, regulating everything from taxes to consumer safety. Some of those laws are set to go into effect early next year. But the Administration says the states are creating a confusing patchwork of rules for AI companies to follow.

The President’s special advisor on AI and Crypto David Sacks joined Trump for a press conference and said the administration will work with members of Congress to draft and pass a national policy for AI regulation.

SACKS: In the meantime, this EO gives your administration tools to push back on the most onerous and excessive state regulations.

But what counts as excessive is not clear in the order.

HODGES: The executive order doesn't provide a national standard…

Wes Hodges leads the Center for Technology and the Human Person at the Heritage Foundation.

HODGES: …What it does is it gives a a vision for what the administration wants, which is a, you know, minimally burdensome, regulatory regime to govern AI at the federal level.

Earlier this year, the administration lobbied Congress to pass a ten-year moratorium on state-level AI regulations with the aim of freeing AI companies to compete with China at a faster pace. When it came time to pass the president’s big beautiful reconciliation bill:

PRESIDING OFFICER: On this vote, the yeas are 99, the nays are one. The amendment is agreed to.

The AI rule was amended straight into the waste bin. And then this month, efforts to pass a federal AI policy in the annual defense authorization bill also faltered. Now this executive order is putting Congress—and states—on notice:

MORELL: They're concerned about certain state AI laws in particular, and wanted to get in front of them…while a federal standard is being put together.

Clare Morell is a fellow at the Ethics and Public Policy Center, researching tech policy. She says the Administration is not aiming to challenge all state laws regulating AI, but is likely to focus on a particular kind of law.

MORELL: The concern all along has been state laws coming out of California and Colorado that push kind of censorship, or almost like DEI standards through for AI models.

In September, Colorado’s governor signed an updated version of the state’s AI law. It requires companies to protect consumers from what the government calls “risks of algorithmic discrimination.” Enforcement is set to begin in February. Trump’s order calls out Colorado for putting companies in a position of producing false results to avoid differential impact of protected groups.

Meanwhile, this fall, California began regulating AI used in hiring. Morell says the Trump Administration is targeting those laws…and pressuring other states to drop similar bills. But even if the DOJ moves fast to create the litigation task force, overruling the states won’t be easy.

MORELL: It really does take time to prove a legal case and win legal arguments. So these laws in the States, you know whether you like them or not, even these kind of problematic AI laws are not going to go away overnight.

The executive order does lay out another tool for pressuring the states: federal funding for internet broadband projects.

MORELL: They would withhold funding under the broadband equity access and deployment program, which has about $42 billion in funding. So this is a huge program, and a lot of states, especially rural states, really rely on it for their broadband funding.

For now, it’s up to Congress to decide if a federal standard for AI safety and development regulations is necessary. But Wes Hodges at the Heritage Foundation is watching the White House with concern.

HODGES: It would do well to target things that are truly unlawful, which I hope is the intent. But if folks use as a means to strike down, you know, all AI laws as an extreme, that would be too far. And ultimately, not good for building the public will, the good trust that you need to scale a technology in a free society. Ultimately, we want to win here. But the question is, you know, to what extreme are we going to take that language?

For WORLD, I'm Mary Muncy. Washington Producer Harrison Watters reported this story.

WORLD Radio transcripts are created on a rush deadline. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of WORLD Radio programming is the audio record.