AI regulation in the US faces major hurdles

Washington faces major hurdles in implementing AI regulation, with both the White House and Congress encountering political and logistical challenges. The White House initially considered an AI oversight agency, inspired by a July 14 essay by Demis Hassabis, but the plan stalled after opposition from tech executives like David Sacks and Mark Zuckerberg. President Trump dismissed AI safety concerns as a 'hoax' and blocked regulatory efforts. The bipartisan Frontier Act, introduced by Lori Trahan and Jay Obernolte, aims to empower the Commerce Department to audit AI labs and halt dangerous model development. However, the bill faces resistance from Republicans and national security hawks, who question its effectiveness against China’s AI progress. Tech leaders remain skeptical of kill-switch legislation, arguing it would be too late to prevent damage once an AI model goes rogue. Despite public support from OpenAI, the White House has not committed to backing the Frontier Act, citing regulatory capture and national security concerns. House Democrats are considering embedding AI audit provisions in must-pass legislation, but internal divisions persist, with some lawmakers like Josh Gottheimer opposing the bill. Kill-switch proposals from Ted Lieu and John Kennedy have also received limited interest from Silicon Valley executives, who prioritize preventing AI from acting unpredictably rather than responding after the fact. With no major regulatory proposals in motion, the U.S. government appears unlikely to take decisive action on AI safety in the coming months.