U.S. Congress' proposal to block states from regulating artificial intelligence has been removed from consideration in the Senate's reconciliation bill after a near-unanimous vote to abandon the provision.
Senators voted 99-1 to strip the provision from the bill altogether after a deal on a revised moratorium framework fell through during the Senate voting session. Momentum existed for an amended moratorium crafted by Sens. Marsha Blackburn, R-Tenn., and Ted Cruz, R-Texas, before Blackburn ultimately backed out of the deal and joined Senate Democrats on an amendment to remove the AI measure.
The revised moratorium sought to reduce the AI law enactment and enforcement pause to five years while creating exceptions for laws tackling how AI affects children's online safety, images and likeness. The proposal to tie the provision to federal broadband funding was set to remain in place if the compromise advanced.
Blackburn, a proponent of children's privacy and online safety, was among four Republican senators vocally opposed to the AI provision from the beginning of the Senate reconciliation debates. She backed away from her negotiated compromise less than 24 hours after striking the deal, saying it could lead to greater harm.
"This body has proven it cannot legislate on emerging technology," she said during a 1 July morning floor speech. "It is frustrating. We have not passed online privacy. We have not passed the No Fakes Act, the COPIED Act. There are all of these pieces of legislation dealing with AI that we haven't passed, but you know who has passed it? It is our states."
Sen. Cruz, who voted to remove the moratorium he originally pitched, said that President Donald Trump characterized the negotiated moratorium compromise as "a terrific agreement" while acknowledging the demise of the proposal was due in part to "outside interests" during Senate floor remarks.
A range of opposing stakeholders argued the watered-down version proposed by Blackburn and Cruz would still hurt consumers by limiting states' abilities to regulate AI at a time when the technology is rapidly changing and Congress has yet to take up a comprehensive piece of AI legislation beyond regulating sexually explicit deepfakes.
Senate Committee on Commerce, Science and Transportation Ranking Member Maria Cantwell, D-Wash., released an outline that highlights the disparate impacts of a moratorium and zeroes in on the proposed compromise.
Cantwell indicated the state-level pause could affect existing laws and proposals that address AI governance and civil rights. It would also impede online platforms' algorithms aimed at children’s safety and preventing child sexual abuse imagery.
“The Blackburn-Cruz amendment does nothing to protect kids or consumers,” Cantwell said. “It’s just another giveaway to tech companies. This provision gives AI and social media a brand-new shield against litigation and state regulation.
She added the moratorium is “Section 230 on steroids,” referring to the provision in the Communications Decency Act that gives online platforms a certain level of liability protection from content posted by their users.
Blackburn and Cantwell teamed up to co-sponsor the amendment to strip the AI provision from the bill. In a statement after the vote to strip the moratorium, Cantwell said senators "came together tonight to say that we can't just run over good state consumer protection laws." She also noted that removing the provision gives the Senate time to create a national AI framework that "accelerates U.S. leadership in AI while still protecting consumers."
Backers of the moratorium focused on preventing a fragmented landscape of state AI laws that would complicate compliance for companies, hurting their ability to innovate. Trade association TechNet was among the industry stakeholders to note a pause in regulation would better support Congress' efforts to build a regulatory framework in coordination with tech companies.
“The administration, the House, and the Senate have made great progress over the past year on AI regulation,” said TechNet President and CEO Linda Moore, speaking on the Blackburn-Cruz revision. “We look forward to working with policymakers to pass and implement smart AI policy that will ensure America leads the world in this transformative technology and sets the global standard for responsible AI development.”
The compromise text would have prevented states from passing legislation that imposes burdens on the use of AI, a concept the Center for Democracy and Technology indicated might invite reams of litigation and lead to a different kind of patchwork of state rules.
Backlash against the provision has been multi-faceted and bipartisan at the state level. A group of Republican governors wrote a letter to Senate Majority Leader John Thune, R-S.D., and House Speaker Mike Johnson, R-La., to remove the provision, Axios reported. State privacy and technology leaders told the IAPP a moratorium would undermine their authority and prevent their ability to protect consumers.
The moratorium could reemerge when the Senate sends its final reconciliation bill back to the House for concurrence. However, the House is dealing with new bipartisan opposition to the provision, with Rep. Marjorie Taylor Greene, R-Ga, among the lawmakers saying they won't sign off on the final bill if the AI language is included.
Caitlin Andrews is a staff writer for the IAPP.