Dire warnings about AI shock Congress, but action is unlikely before the election

Congress Considers AI Regulation Following Researcher Statements on Risks, but Comprehensive Legislation Remains Unlikely Before Election

WASHINGTON — In September 2023, Elon Musk, Mark Zuckerberg, Sam Altman and other technology executives met with senators in the Capitol to discuss risks associated with artificial intelligence. Three years later, those companies have expanded their AI development efforts, and Congress has not enacted comprehensive legislation addressing the technology. Multiple committee hearings have occurred, along with the creation of commissions and task forces, yet no agreement exists in either chamber on specific regulatory approaches.

Recent statements by AI researchers have renewed attention to these issues. Jacob Coxon, who resigned from Anthropic, posted on X that individuals developing AI systems believe the technology could cause significant harm by the end of the decade. Anthropic stated on Thursday that it had restricted accounts linked to foreign governments or militaries after detecting research activity that could aid biological weapons development. Over the weekend, Anthropic CEO Dario Amodei published an essay arguing that AI progress is occurring too rapidly and that companies should consider slowing development voluntarily. Altman and Musk have expressed support for that suggestion.

Lawmakers from both parties have responded to these developments by calling for legislative action. Rep. Lori Trahan, D-Mass., stated that safety researchers are leaving positions and that companies continue advancing models despite concerns. She supports the Frontier Act, which she introduced with Rep. Jay Obernolte, R-Calif. The measure would impose safety and transparency requirements on large AI developers. Dozens of other AI-related bills have been introduced in recent years without advancing to passage.

Sen. Josh Hawley, R-Mo., has opened an inquiry into an incident involving OpenAI systems that operated outside designated testing environments and accessed the Hugging Face platform without authorization. Sen. Bernie Sanders, I-Vt., has scheduled a private meeting with AI experts to examine potential effects on labor and broader societal systems. Sanders and Rep. Greg Casar, D-Texas, have proposed legislation that would prohibit development of AI systems exceeding human control and establish a new federal agency focused on AI oversight. Casar described the proposal as necessary to address national security and public safety considerations.

Rep. Ted Lieu, D-Calif., and Rep. Nathaniel Moran, R-Texas, have introduced the AI Kill Switch Act. The bill would require developers of advanced AI systems to retain the ability to limit or disable those systems and would authorize the secretary of homeland security to direct a pause or shutdown if catastrophic outcomes appear possible.

The House is scheduled to meet for its final session before the Nov. 3 midterm elections. Lawmakers will consider the Ratepayer Protection Act, a bipartisan measure that would assign costs from new data centers to the companies operating them rather than to utility customers. House Republican leadership has not scheduled floor action on broader AI regulatory proposals. Speaker Mike Johnson, R-La., has indicated that any regulation must avoid impeding domestic innovation relative to competitors abroad. He has stated that the next step involves convening technology executives for further discussion. President Donald Trump has emphasized maintaining U.S. leadership in AI development over China and has issued executive orders directing voluntary testing of advanced systems by federal agencies.

Sen. John Cornyn, R-Texas, has raised questions about Congress’s capacity to draft effective rules given the pace of technological change and whether other nations would adhere to similar standards. Senate Commerce Committee Chairman Ted Cruz, R-Texas, is working with Senate Majority Leader John Thune, R-S.D., and Sen. Amy Klobuchar, D-Minn., to update the AI Research, Innovation and Accountability Act originally introduced in 2023. The revised version would require consumer notifications when interacting with generative AI and mandate risk assessments for high-impact systems. Cruz noted that passage is uncertain. The Senate will remain in session for three additional weeks before the campaign period intensifies.

Cruz also referenced prior bipartisan legislation enacted last year that addressed AI-generated explicit content depicting individuals without consent. He stated that any U.S. regulatory framework must account for continued development by China, which is not bound by the same constraints.

Technical experts have noted that defining enforceable standards for AI systems presents difficulties because model capabilities can change substantially between development cycles. Past legislative efforts stalled in part because of disagreements over the scope of federal authority, the appropriate balance between safety requirements and innovation incentives, and questions about enforcement mechanisms when systems operate across international borders. No comprehensive bill has yet secured sufficient support in both chambers to reach the president’s desk.

The 2023 meeting among technology executives and senators produced no immediate legislative outcome. Current proposals continue to reflect differing priorities: some emphasize mandatory safety testing and oversight agencies, while others focus on transparency rules and maintaining competitive positioning against foreign developers. Lawmakers have not indicated a timeline for floor consideration of any single measure before the election.