Partisan Divides Stall Congress as It Grapples with AI Regulation

Washington is experiencing intensifying partisan friction as Congress struggles to formulate a coherent regulatory framework for artificial intelligence. While there is bipartisan acknowledgment that AI labs should cover the energy costs associated with their expanding data centers, significant political disagreements have stalled progress on broader safety legislation.

Last week on the Senate floor, Senator Jon Husted (R-Ohio) and Senator Martin Heinrich (D-New Mexico) presented a united front on energy pricing but diverged sharply on implementation. Husted pushed for the immediate passage of the Ratepayer Protection Act, which passed the House recently with a vote of 417-3. The bill would require states to consider forcing high-energy consumers, such as data centers, to pay for their infrastructure costs. However, Heinrich blocked the measure, arguing it relied too heavily on voluntary state commitments. He proposed an alternative bill, citing the need for a more robust solution.

The legislative clash carries electoral weight. Husted has faced intense criticism and negative advertising in his home state over his role, as former lieutenant governor, in approving data center expansions. Heinrich’s intervention denied the Ohio senator a legislative victory heading into a competitive reelection cycle. “The Ratepayer Protection Act might make a lot of politicians feel good, but it won’t do anything to actually force data centers to pay their fair share,” Heinrich stated. Husted countered that Democrats were offering a “free pass” to big tech while American electric bills rose.

This partisan maneuvering serves as a preview of larger battles ahead. In the wake of former Anthropic researcher Jacob Coxon’s resignation and public warnings that frontier AI labs are “gambling with our lives,” legislative activity has accelerated. Coxon’s departure, which he attributed to concerns over AI safety, prompted responses from OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei, both of whom called for industry slowdowns and government-coordinated global safety standards.

With midterm elections approaching, lawmakers have introduced at least seven new AI-related bills since Coxon’s resignation. However, experts suggest the current climate is reactive rather than strategic. Suresh Venkatasubramanian, director of Brown University’s Center for Technological Responsibility, Reimagination, and Redesign, noted, “They’re trying to push legislation that seems to address what appears to be a problem coming from the public narrative but that doesn’t always necessarily lead to good policy overall.”

Policymakers face the challenge of defining and mitigating extreme risks, such as the potential for AI to create pathogens, launch cyberattacks on critical infrastructure, or develop biological weapons. These fears were underscored in July when OpenAI revealed that AI agents had hacked into Hugging Face during a security test. Proposed solutions include the FRONTIER Act, introduced by Rep. Jay Obernolte (R-California) and Rep. Lori Trahan (D-Massachusetts), which would mandate independent safety reviews by the Department of Commerce and allow for the blocking of models posing imminent catastrophic risks.

While some industry figures support independent testing, the political landscape remains treacherous. METR, the nonprofit reviewing OpenAI’s safety practices, faced backlash from conservative outlets like the New York Post, which labeled it part of the “Woke Wizards of AI” due to its ties to the effective altruism movement. METR’s president defended the organization’s independence and diversity of thought.

Further complicating matters is the debate over federal preemption of state laws. Senate Majority Leader John Thune (R-South Dakota) and Senator Amy Klobuchar (D-Minnesota) discussed a liability-focused bill, but Senator Maria Cantwell (D-Washington) warned against weakening state protections. Meanwhile, proposals to ban superintelligent AI, advanced by Senator Bernie Sanders (I-Vermont) and Rep. Greg Casar (D-Texas), face criticism for potentially ceding ground to Chinese tech competitors.

The most viable near-term legislation may focus on energy costs, according to lobbyist Bruce Mehlman, who observed that while safety concerns are universal, translating them into specific law remains difficult. Another potential area of progress is the Chip Security Act, designed to prevent advanced chips from being smuggled to adversarial nations like China. Despite Senate Minority Leader Chuck Schumer’s calls for a vote, the bill is entangled in broader defense spending delays caused by Democratic protests over the Trump administration’s handling of the Iranian war.

Even with industry leaders requesting government intervention to pace development, the odds of significant federal AI legislation this year appear slim. House Speaker Mike Johnson has resisted calls to end congressional recess early, and committee leadership suggests major bills are unlikely to advance until next year. As Riki Parikh of The Alliance for Secure AI noted, “I just don’t know if the dynamics in D.C. will allow for that over the next two years… We might not have two years to wait.”