AI leaders split with US politicians over calls to slow frontier development

Anthropic chief executive Dario Amodei has called for frontier artificial intelligence development to be paced, arguing that companies should take more time to align and safeguard their models. His proposals include independent third-party safety evaluators, coordination among leading AI companies on standards and limits, and international government cooperation.

OpenAI chief executive Sam Altman endorsed the approach and said his company would support independent evaluators. He also backed a consistent federal framework for frontier AI safety requirements, while stressing that pacing meant progress should be slower than it otherwise might be, not halted.

Google DeepMind co-founder Demis Hassabis said Amodei’s essay pointed in the right direction, although he said the details still needed to be worked through. Elon Musk also expressed support for Amodei’s position.

US President Donald Trump rejected the idea of slowing AI development, saying strong presidential oversight and existing criminal and regulatory powers were sufficient. He also framed the issue as a competition with China, arguing that the country that wins AI will lead.

Vice-president JD Vance questioned why AI companies were seeking government regulation, while House Speaker Mike Johnson called for discussions about balancing model safety with national security and competition with China. The source institution, The Verge, reported that Johnson suggested a meeting with AI companies but that it was unclear whether or when it would take place.

Senator Bernie Sanders supported a call to pause AI development. The competing positions underscore a widening dispute over whether safety safeguards should slow the frontier or whether doing so could undermine national security and technological competition.