White House scrambles to tame AI fears

The White House is facing internal divisions over AI regulation as reports emerge of a potential executive order mandating government vetting of new AI models before public release. The shift comes after Anthropic’s latest model, Mythos, demonstrated the ability to identify decades-old security vulnerabilities, prompting concerns about unchecked AI development. Earlier this week, The New York Times reported the administration was exploring vetting requirements for AI models, while Politico later detailed plans for a formal approval process resembling an FDA-like system. National Economic Council Director Kevin Hassett suggested an executive order could establish a clear framework for evaluating AI risks before release, comparing it to drug approval processes. The proposed measures have drawn sharp criticism from the tech industry, which argues they align more with the Biden administration’s cautious approach than Trump’s deregulatory priorities. A former Trump White House official criticized the administration’s inconsistent messaging, calling it ‘flip-flopping’ and distracting from urgent AI risk mitigation efforts. Industry players warn that strict vetting could stifle innovation, while critics of overregulation argue the shift reflects broader uncertainty in the administration’s tech policy. The back-and-forth has intensified debates over balancing AI safety with maintaining U.S. competitive advantages in the global market.