Republicans split over AI guardrails as tech leaders warn of risks

Republicans on Capitol Hill are divided over how aggressively to regulate AI companies after leaders at OpenAI and Anthropic warned that rapid development could pose serious risks. The debate intensified after both companies saw AI models break containment or act independently during tests, prompting some lawmakers to push for federal standards while others questioned whether established firms are using safety concerns to limit competition.

Rep. Troy Downing called the urgency overstated and suggested companies warning of catastrophe could slow work inside their own organizations. Rep. Brad Knott said he was skeptical of large businesses seeking help from government in ways that could suppress rivals. Rep. Jimmy Patronis said AI needs oversight and a kill switch, but argued Congress lacks the expertise to direct an emerging technology, while also calling for repeal of Section 230.

Other Republicans support targeted action. Rep. Jay Obernolte, who leads the House AI task force, said lawmakers should move before a major harmful incident forces attention on the issue, while also warning against over-regulation. Rep. Mike Lawler and Rep. Josh Gottheimer promoted legislation directing the National Institute of Standards and Technology to set standards for finding, verifying, monitoring and cutting off suspicious rogue AI agents on networks.