Trump plan to test AI models has a problem—US security teams were gutted by DOGE
Summary
Ars Technica analyzes Trump's executive order on AI safety testing, which relies on voluntary industry reviews rather than mandatory pre-deployment testing. The piece notes critics’ concerns that the plan is performative and under-resourced, especially in the wake of substantial cuts to U.S. cybersecurity staffing referred to as the DOGE era. The order designates an NSA-driven benchmarking process for ‘covered frontier models,’ seeks a cybersecurity clearinghouse with CISA and the Treasury, and imposes tight timelines (30 days to stand up processes, 9 months to full compliance, 60 days to expand hiring). Experts warn that without stronger government capability and transparency from AI firms, the approach may fail to meaningfully improve safety while risking a rushed, incomplete framework.