President Trump signed an executive order on June 2 directing the federal government to establish a voluntary early review process for America’s most advanced AI models — asking companies to allow government agencies, including the NSA and the Defense Department, to evaluate cutting-edge models up to 30 days before public release.

This is a reversal. Less than two weeks earlier, Trump scuttled an earlier version of the policy that gave the government a 90-day review period. The shorter 30-day window — pitched by former AI advisor David Sacks as a “game changer” that lets companies engage with government without slowing down releases — represents a compromise. But the key word is “voluntary.” The order explicitly stops short of mandating participation and bars the creation of any new licensing or permitting for AI models.

The order formalises an existing practice: top AI companies already share models with external evaluators and government players before deploying them publicly. What’s new is the institutional framework — a explicit, timestamped process with national security agencies at the table.

But here’s the tension. The administration came to power promising to strip away AI guardrails. Now it’s asking the industry to voluntarily hand over their most valuable intellectual property — frontier models that represent billions of dollars in R&D — for government review before anyone else can see them. That’s a hard ask without a mandate.

“The main question is whether this is the start of a continued government clamp down and response to continued AI capabilities, or whether this is a one-off, limited, and truly voluntary act,” said James Sanders at the Center for a New American Security.

The next six months will tell. If major labs participate willingly, this becomes a model for government-industry cooperation. If they don’t, the order becomes a paper tiger — and the push for mandatory review returns.