The U.S. House of Representatives adjourned on Wednesday, effectively postponing any legislative action on artificial intelligence regulation until after November’s midterm elections. The decision comes despite intensifying pressure from AI industry leaders and some members of Congress to address safety concerns following claims by an Anthropic researcher that the technology could pose existential risks by the end of the decade.
Speaker Mike Johnson, Republican of Louisiana, announced the early recess to allow lawmakers to return to their districts and campaign ahead of the midterms, which will determine control of Washington for the remainder of President Donald Trump’s term. With early voting already underway in several states, Johnson emphasized the political necessity of the move.
“It’s time for our members to go home to their districts and make their case to the American people,” Johnson told reporters on Wednesday, confirming he had canceled the House’s scheduled work for Thursday. He indicated that the responsibility for passing AI legislation now falls to the Senate, which plans to remain in session for approximately two more weeks. For any bill to become law, it must pass both chambers and be signed by the president.
The adjournment runs counter to the positions of Democrats and a faction of Republicans who argue that Congress cannot afford to wait. Rep. Sam Liccardo, a Democrat from California representing part of Silicon Valley, led 107 Democratic colleagues in a Tuesday letter to Johnson demanding the House remain in session to enact AI safeguards. The letter criticized Congress for “fiddling” while risks escalate, noting that floor time was recently consumed by debates on resolutions opposing socialism before the six-week recess begins.
“Our children will not forgive us for our inaction,” Liccardo said, invoking the historical metaphor of Nero fiddling while Rome burned. He called for immediate action on what he described as “low-hanging fruit,” specifically pointing to the Frontier Act. Sponsored by Rep. Jay Obernolte, Republican of California, and Rep. Lori Trahan, Democrat of Massachusetts, the bill would mandate third-party safety audits for AI labs, introduce transparency requirements, and grant the Commerce Department authority to suspend models posing an “imminent catastrophic risk.”
Liccardo also suggested Congress consider a “kill switch” provision to disable rogue AI models and explore a narrow antitrust exemption allowing top tech companies to collaborate on safety issues without facing legal repercussions.
House Speaker Johnson has previously cautioned against rushing AI regulation, warning that hasty legislation could disadvantage the United States in the global competition with China. President Trump has similarly opposed restrictive measures, dismissing recent safety warnings as a “hoax.”
Despite the House’s departure, Rep. Obernolte stated that the urgency of the situation requires congressional action before the year ends. He plans to push for the Frontier Act and the Great American Artificial Intelligence Act, a comprehensive regulatory framework he developed with Trahan, to be considered together when lawmakers return. However, Obernolte acknowledged that meaningful regulation before the election was unlikely due to political uncertainties.
In the Senate, Commerce Committee Chair Ted Cruz, Republican of Texas, along with Senator Amy Klobuchar, Democrat of Minnesota, and Senate Majority Leader John Thune, Republican of South Dakota, are drafting an AI safety bill. Cruz hinted this week that a committee vote could occur by the end of September, though he noted that bipartisan agreement remains a critical question.
Cruz drafting a bill is a funny sight. I doubt we will see real safety audits before November, regardless of what the headlines say.
Typical. They only care about safety when it helps their campaigns. The AI risks won’t wait for election season to materialize.