Calls for federal rules on artificial intelligence are colliding with political realities in Washington. President Trump dismissed warnings about AI risks and signaled opposition to new guardrails, narrowing prospects for near-term legislation in a Republican-led Congress. House Speaker Mike Johnson said companies should prioritize self-regulation and hinted only at a potential, yet undefined, congressional role. With the House poised to recess before the midterms, Democratic leaders urged delaying the break to consider safeguards, but movement this week appears unlikely.
Lawmakers are divided not just along party lines but over policy design. Competing proposals range from stricter oversight to a kill switch and even a moratorium on data centers. In the Senate, Republicans’ John Thune and Democrat Amy Klobuchar are drafting a bill to subject advanced AI firms to federal oversight, while in the House, Democrat Lori Trahan and Republican Jay Obernolte are pushing the Frontier Act to empower the government to halt risky models and require independent audits. Meanwhile, Sen. Bernie Sanders is pressing for a sweeping pause on infrastructure growth.
Industry voices are split. Venture capitalist David Sacks downplayed existential risks and urged company-led safety commitments, while Anthropic’s Dario Amodei has called for slowing frontier development. Policy experts say the administration’s voluntary assessment framework lacks transparency, and they doubt Congress will pass comprehensive AI legislation soon. For now, companies’ self-imposed safety regimes may fill the vacuum as Washington debates how—and whether—to regulate.
Related articles:
NIST AI Risk Management Framework
OECD AI Principles
California SB 1047: Safe and Secure Innovation for Frontier Artificial Intelligence Models
