White House Limits National Security Testing Framework for Artificial Intelligence Models
The Trump administration’s artificial intelligence safety framework will focus on closed-source frontier models, prompting debate over how regulators should address security risks from open-source and open-weight systems.