Analysts urge US to unilaterally pace frontier AI development

Policy analysts propose that the United States government immediately implement unilateral restrictions on frontier AI development to mitigate risks associated with full R&D automation. A post by researcher Romeo Dean argues that existing authorities allow for immediate pacing without significant legislative preparation. Proponents contend that slowing domestic progress reduces safety risks during critical transition periods and limits technology diffusion to competitors like China through distillation or theft. The strategy aims to build regulatory capacity and create leverage for future international agreements while preserving US leadership. However, the proposal acknowledges significant geopolitical risks, including prematurely surrendering technological advantages during lower-risk windows. Analysts suggest that establishing domestic controls now creates necessary infrastructure for durable global governance later. This approach prioritizes long-term safety coordination over short-term competitive maximization as autonomous AI research capabilities emerge.