Trump executive order asks frontier labs to share new models with government for up to 30 days — voluntary, but tied to "trusted partner" early-access designations
President Trump signed an executive order on June 2 that asks technology companies to voluntarily share new AI models with the federal government for up to 30 days before public release, and to collaborate with the administration to select "trusted partners" who gain early access. The order signals a structural shift from the administration's prior hands-off posture toward frontier oversight.
The voluntary framing is the lever. The EO does not impose mandatory pre-release disclosure — that would clash with the administration's broader deregulation posture and the recently-stalled Colorado AI Act enforcement. Instead, it bundles voluntary disclosure with the upside of being designated a "trusted partner" eligible for federal procurement, critical-infrastructure cybersecurity contracts, and early-access privilege programs. Labs that decline disclosure are not penalized; labs that participate gain access to a market segment Anthropic's Project Glasswing has demonstrated is real and lucrative.
The strategic read is closer to Project Glasswing than to the EU AI Act. Anthropic, OpenAI, and Google already run informal early-access programs for federal customers and have already accepted disclosure as the cost of accessing classified-system contracts; the EO formalizes that pattern and makes the federal procurement piece the explicit incentive. The administration gets 30-day pre-public visibility into frontier capability; the labs get a structured path to defense-tier revenue. Voluntary on paper, mandatory in any lab's procurement-driven calculus.
White House — Promoting Advanced Artificial Intelligence Innovation and Security → · Scientific American — Trump's new AI executive order drastically shifts the administration's stance on the tech →