Trump Signs Executive Order For AI Testing Prior To Frontier Model Releases
Zvi's AI Roundups · Zvi Mowshowitz · 2026-06-03
Zvi Mowshowitz analyzes Trump's newly signed AI executive order—creating voluntary 30-day pre-release government access for frontier models—arguing it functions as a de facto licensing regime despite explicit disclaimers, with NSA involvement and classified benchmarks posing significant abuse risks.
Appears in
Extraction
Topics: ai-regulationus-ai-governancefrontier-ai-policyai-safety-policyexecutive-order
Claims
- The executive order establishes a voluntary framework giving the US government up to 30 days of pre-release access to frontier AI models, reduced from an earlier draft's 90-day window.
- Despite explicitly disclaiming to create a licensing regime, the EO functions as a de facto mandatory licensing and prior restraint mechanism because labs have no real choice but to participate.
- Classifying the benchmarking thresholds is particularly problematic because researchers at AI labs will not know whether their own models are subject to the review requirement.
- The NSA's central role in administering evaluations is a mistake; the civilian agency CAISI would be a more durable and trustworthy administrator.
- The EO represents a political victory for safety-oriented White House factions over David Sacks and the no-regulation wing, despite Sacks framing the 30-day concession as a win.
Key quotes
Make no mistake. This is a de facto mandatory governmental licensing, preclearance and permitting requirement. Welcome to The Prior Restraint Era.
The fact that the administration is classifying the details of how this 'voluntary' system will work is egregious. The public and the employees of the labs have a right to know how this works.
We had our opportunity to do 'this is hard to abuse' forms of frontier model regulation, to do things the easy and better way... That window is closed now. We have to make policy within the Trump White House.