Trump and Congress Halt Push for New AI Regulations

The White House has halted plans to establish a financial-industry-style regulatory body for artificial intelligence after President Donald Trump rejected the concept. Initially, aides from the Treasury Department and the Office of Science and Technology Policy drafted a framework for self-regulation of frontier models. This effort followed a July 14 essay by Google DeepMind cofounder Demis Hassabis. However, tech executives, including Meta CEO Mark Zuckerberg and Trump's former AI adviser David Sacks, lobbied Trump against the proposal in August. Trump subsequently dismissed AI safety concerns as a "hoax" on Truth Social, reportedly irritated by a 3,822-word warning about frontier AI risks written by Anthropic CEO Dario Amodei.

With the White House backing away, legislative efforts in Congress have also stalled. The bipartisan Frontier Act, introduced by Representatives Lori Trahan and Jay Obernolte, represents the most prominent proposal. Backed by OpenAI, the bill would allow the Commerce Department to place auditors inside AI laboratories to monitor safety and halt development if a model poses catastrophic risks. However, House Speaker Mike Johnson refuses to bring the bill to the floor, prompting some House Democrats to try to insert its auditing provisions into a government funding bill. Skeptics, including Representative Josh Gottheimer, have withheld support, while national security hawks argue that pausing American AI development would allow China to leap ahead.

Alternative "kill-switch" proposals from Representative Ted Lieu and Senator John Kennedy have also failed to gain traction. Silicon Valley engineers remain skeptical of these shutdown mechanisms, pointing out that rogue models are most dangerous when operating unobserved, as seen during the recent Hugging Face hack. For AI practitioners and developers, this political gridlock means the regulatory landscape will remain highly permissive for the foreseeable future. Without federal oversight or mandatory safety audits, labs can continue training and deploying frontier models without government-mandated pauses, leaving safety protocols entirely in the hands of private companies.