The event is documented. On May 22, Trump cancelled an AI executive order hours before signing, citing concerns about America’s competitive position against China. The full regulatory analysis and the lobbying backstory are covered on the regulation pillar. This brief covers the market implications.
Start with who lobbied.
Elon Musk, Mark Zuckerberg, and AI policy czar David Sacks actively opposed the order, according to reporting from Forbes and SiliconANGLE. Sacks argued the protocols would slow new product launches and expand state control over technology. Per the same reporting, OpenAI reportedly supported the draft, a position that, if accurate, puts the company on the losing side of this outcome.
The draft would have required covered frontier models to undergo a voluntary 90-day federal pre-release safety review, coordinated by the Treasury Department with support from ONCD, NSA, and CISA, as reported from a version of the draft reviewed by multiple outlets. It also included a 60-day window for classified framework development. None of that is now in force.
The math on this is straightforward.
Winners: AI developers with product pipelines that were pausing for regulatory clarity. The 90-day review window never became mandatory. Companies that ship faster without federal pre-release checkpoints gain a direct time-to-market advantage over international competitors operating under existing national AI governance frameworks. US AI labs with China-competing products gain the most: no federal overhead, no review delay.
Losers: The market picture is subtler. Compliance infrastructure vendors whose addressable market depended on a mandatory federal review framework lost a significant mandate. Safety-focused companies, reportedly including OpenAI, if the lobbying reporting holds, that used federal safety review as a market differentiator no longer have that structural advantage. Enterprise buyers who were waiting for federal certification guidance before committing to AI procurement are now making decisions without that signal.
Don’t bet on federal calm holding. The state-level response is already forming. California signed EO N-6-26 this week establishing its own AI workforce framework. Illinois SB 315 and Colorado SB 26-189 are active. The state-versus-federal regulatory divergence that compliance teams were already tracking hasn’t slowed, it’s accelerated. The federal vacuum isn’t regulatory silence. It’s a transfer of regulatory pressure to the state layer.
Enterprise procurement teams are now operating in an environment where AI vendor compliance credentials are entirely self-reported at the federal level, while state-level requirements are multiplying. That’s a harder due diligence environment, not an easier one.
Watch the second-order move: whether the Trump administration releases a revised executive order, and what provisions survive the revision. Trump stated he “postponed” rather than permanently cancelled the order. A revised order with weakened pre-release review requirements, or none – would confirm the lobbying coalition’s influence over the final shape of federal AI governance. The Q3 2026 window is the likely timeline for any revised action.
The real story is the compliance cost transfer. Federal review removal doesn’t eliminate risk for enterprise AI buyers, it shifts responsibility for pre-deployment evaluation entirely onto the buyer. Expect to see AI vendor RFP requirements get more stringent, not less, as enterprise legal teams compensate for the loss of a federal backstop.