Trump wants the gridlocked Congress to move forward on AI. Frustrated states say they already have: NPR

In the absence of action at the federal level, states are passing dozens of laws of their own to regulate artificial intelligence — creating guidelines for child safety, requiring greater transparency from the technology and ensuring whistleblower protections.

But state lawmakers — including those from President Trump’s party — are facing pushback from the White House. Trump and his advisers, including AI and crypto mogul David Sacks, have argued that various state laws are a burden to innovation.

“We want to create an environment where innovators have certainty about how they can develop their products, and that’s something only Congress can provide,” Michael Kratsios, head of the White House Office of Science and Technology Policy, said earlier this week. “The first step is to create a national framework so we can avoid patchwork.”

The White House recently issued a Regulatory framework for artificial intelligence That the administration wants Congress to enact an Act.

The “patchwork” of state laws is something that administrations — including the president himself — have repeatedly criticized, so much so that they have gotten involved in the work that some states are doing on AI, even though it is led by Republicans.

State Rep. Doug Fifia, R-Utah, proposed a bill in his state’s legislature earlier this year. The aim was to require technology companies to be more transparent about how they will protect consumers.

But the bill never even came up for a vote — thanks to interference from the Trump administration that came in the form of a one-line memo, he said.

“They basically mentioned that they opposed the bill and did not believe the bill was reformable and that it was against the administration’s AI agenda,” Fifia said.

No explanation was given as to what this meant. A White House official, speaking on background because they are not authorized to do so publicly, told NPR that the White House has never told any state it cannot enforce child protection protections, though he did not comment specifically on the memo sent to Fife.

Fifia, a former Google employee, says the memo wasn’t “a huge shock” because he was already hearing from the Trump administration about concerns about his bill. And he said state and federal lawmakers should be involved in the issue of AI regulation — but there are limits to what Congress can do.

“Congress is at an impasse and they not only won’t act, they can’t act. In states like Utah we see this as an opportunity to move forward and protect our constituents and citizens, especially as it relates to child safety,” Fifia said.

similar approach

Other Republican state lawmakers are taking the same approach.

Pennsylvania Republican state senator Tracy Pennycuik said, “I have little interest in what the federal government is doing at this point. I mean, I’m sorry. It takes too long.” “I think when there’s a problem the states are the first to see and have the ability to act quickly. So we’re going to continue what we’re doing.”

Pennycuik has sponsored legislation regulating AI, including recent legislation SafeChat Act In the state, AI companies are required to include safeguards to prevent chatbots from providing content that encourages self-harm or violence toward themselves or others.

In Texas, state Senator Angela Paxton, a Republican, is doing similar work.

“We don’t want to have a patchwork of regulatory structures… in general it’s a good rule of thumb,” Paxton said. But she says tech companies don’t have a good history of regulating themselves.

“When you have no regulation, you have the Wild West,” she said. “I like the idea of ​​creating stronger federal laws but until that exists, I think we have to preserve the states’ ability to pass laws.”

The White House outline outlines some of the principles for how the Trump administration wants Congress to move forward on rapidly changing technology, including protecting children from harm and consumers from the rising costs of data centers.

mixed reaction

There has been a mixed response from lawmakers and experts, who say it’s good to have one set of rules, but the White House framework lacks detail.

Ricky Parikh, policy director at the nonprofit Coalition for Secure AI, said the framework is not specific enough on issues such as the technology’s potential role in job replacement. It’s also not enough to hold technology companies accountable.

“A federal standard is better than a 50-state patchwork,” Parikh said. “But what they are proposing here is not enough. It does not give the right to change the good work that is being done by the states.”

Tennessee Attorney General Jonathan Scrametti sees the administration’s framework as a step in a positive direction. He is more optimistic about White House action on this framework than previous steps; He was particularly concerned about the White House’s push last year to impose a 10-year moratorium on state AI laws, a move deemed favorable by technology companies.

“If you had enforcement authority and a 10-year lockout of anyone who wanted to use it, who knows what would have happened,” Scarmetti said. “That was really scary.”

Efforts by the White House and allies like Republican Senator Ted Cruz of Texas to force that moratorium ultimately failed. Still, Scermetti remains concerned about Trump’s closeness to the AI ​​industry.

This sentiment is also reflected in how the public feels about AI. January 1st survey Polls conducted by Morning Consult and the Tech Oversight Project showed that a majority believe the Trump administration is too close to Big Tech. Recently polled by Vanderbilt University It also shows that Republicans are more in favor of regulating artificial intelligence than Democrats.

On Capitol Hill, the president’s framework has drawn support from his Republican colleagues, but actual movement on legislation is still pending.

Tennessee Republican Senator Marsha Blackburn said she has been in contact with the White House about her Trump America AI Act, a bill that expands extensively on the White House’s four-page framework.

“By releasing a national framework on AI, the Trump administration has given us a roadmap to craft legislation, and now it is Congress’s turn to pass a bill that will codify the president’s agenda, protect Americans, and spur AI innovation,” Blackburn said in a statement.

The White House says it continues to have “productive conversations” with lawmakers.