US Government Expands Oversight of Frontier AI Models

The Commerce Department’s Center for AI Safety and Innovation (CAISI) is spearheading the effort, aiming to conduct pre-deployment assessments and targeted research. This protocol builds on a framework already established with OpenAI and Anthropic, which have undergone 40 separate reviews since 2024. These companies recently renegotiated their agreements to align with the specific priorities outlined in President Donald Trump’s AI Action Plan.

Beyond these current partnerships, the administration is weighing more aggressive regulatory steps. Reports indicate that the White House is considering an executive order designed to formalize collaboration between tech leadership and federal officials for ongoing model oversight. Chris Fall, director of CAISI, emphasized that rigorous, independent measurement remains vital for evaluating national security implications as the industry scales its output.