The postponement of what would have been a major presidential directive on AI and cybersecurity signals that the administration's approach to AI governance remains unsettled—contradicting the narrative that we're moving toward clearer federal oversight.
What We Know
According to reporting from Axios, the executive order was ready for signing Thursday with a ceremonial event planned. But Trump personally pulled the plug after reviewing the draft. Sources suggest the sticking points were around giving the Treasury Department a leading role in AI policy coordination and Trump simply not wanting more regulation on his watch.
"He just hates regulation. That's basically what it comes down to."
This aligns with a pattern we've seen: the administration acknowledges AI as strategically critical but treats prescriptive governance as a political liability. The irony is unavoidable—China, the EU, and other competitors are actively building formal AI strategies while the U.S. relies on market forces and ad-hoc arrangements.
The Bigger Picture
In the same 48-hour window, the Commerce Department unveiled a $2B quantum computing grant program with equity stakes—the kind of industrial policy intervention that the postponed EO apparently avoided. IBM alone is set to receive $1B of that package, and quantum computing stocks surged 30%+ on the news.
Taken together, these moves reveal an administration comfortable spending on next-generation compute infrastructure but reluctant to establish clear rules of the road for AI development and deployment. That's a calculable risk for enterprises building on U.S. AI infrastructure: the lack of a coherent regulatory framework creates uncertainty, even as capital flows freely.
The EO will likely resurface in some form. But the message from the Oval Office is unambiguous—no AI regulation is preferable to the wrong AI regulation, at least for now.