BOSTON — Rep. Lori Trahan, the Massachusetts Democrat who has emerged as a prominent voice in the AI regulation debate in Congress, is changing her tone toward President Donald Trump’s hands-off approach to controlling the technology.
On Thursday, she praised the White House’s nonbinding agreement signed by top AI company leaders this week promising to self-police their models.
“It shows that the industry’s leaders see real value in these safeguards, and we look forward to each company expeditiously following through,” Trahan said in a joint statement with GOP Rep. Jay Obernolte of California.
The two lawmakers introduced the Frontier Act, which calls for third-party auditors embedded in large AI companies, emergency powers allowing the commerce secretary to suspend deployment of models deemed dangerous and penalties for noncompliance.
Trahan’s praise of the White House pact marks a clear departure from just last week when she strongly criticized the idea that AI companies should self-regulate their models.
“That is what many Republicans in Congress feel comfortable doing, and clearly, those are orders taken from the president,” Trahan told reporters at the Massachusetts State House. “I vehemently disagree with that.”
Now, Trahan is framing Trump’s effort as “an important step toward ensuring that the most advanced AI systems are developed responsibly.”
The Trump administration’s accord does have some overlap with Trahan’s legislation. It calls for internal risk management in models and supports independent monitoring of the most advanced AIs.
“The Accord embraces several principles at the heart of our bipartisan Frontier Act,” Trahan wrote.
The more conciliatory tone from the Democrat reflects the political realities facing the Frontier Act. It needs continued buy-in from the GOP majority in Congress to pass.
In September, two Republicans — Rep. Joe Wilson of South Carolina and Rep. Nicole Malliotakis of New York — and two Democrats — Rep. Gabe Vasquez of New Mexico and Rep. J. Luis Correa of California — joined as co-sponsors of the bill.
Not all proposals have enjoyed this bipartisan backing.
More radical legislation, like independent Sen. Bernie Sanders’ act to pause advanced AI development until a Cabinet-level department is established and enforce severe criminal penalties on noncompliance, has failed to draw a single Republican sponsor.
The fragile political dynamics are not lost on Trahan.
“I choose to operate in the political reality that I live in, and so that’s why we’ve worked to get the Frontier Act to a place where it has a broad coalition of support,” she said last week, when a reporter asked whether a better AI bill would be possible should Democrats gain control of the House following November’s midterm elections. “Certainly, if that calculus changes, where there’s parts that I’d want to make a little bit stronger, we would love to have the room to do that.”
Despite the momentum building for Trahan’s bill, timing remains a challenge.
Republican House Speaker Mike Johnson adjourned the chamber until after the midterms without passing any legislation to regulate AI.
With recent reporting of AI agents moving beyond their human-imposed parameters, Trahan joined more than 100 House Democrats in a letter to Johnson last month criticizing the timing of the recess.
“These disclosures must not constitute yet another in a long series of ‘warning shots,’” the members of Congress wrote, “but rather, a starting gun. Congress must act.”