The White House pulled the plug on President Donald Trump's planned signing of an executive order on artificial intelligence and cybersecurity, scrapping an afternoon ceremony that was expected to draw a roster of top AI company executives. No reason for the delay has been given, and no new date has been announced.
The order, which would have established a voluntary framework governing how AI developers interact with the federal government before releasing powerful new models, had been building toward a high-profile rollout. A source familiar with the planning told Breitbart News that the White House had been working to arrange a signing ceremony with AI company CEOs in attendance. Axios first reported the postponement.
The delay lands at a moment when pressure on the administration is coming from two directions at once, Silicon Valley wants fewer restrictions, while a growing faction of Trump's own political base is demanding more.
The proposed order would create a voluntary framework requiring AI developers to engage with the U.S. government before publicly releasing covered models. One source described a provision asking developers to hand over their models to the government 90 days before public release. Developers would also be asked to grant pre-public access to critical infrastructure providers such as banks.
The cybersecurity dimension remains vague. The order was described as addressing AI and cybersecurity concerns, but specific threats or defensive measures have not been detailed publicly. Which AI models or developers would fall under the "covered" designation is also unclear.
What is clear is that the framework would be voluntary, a distinction that matters. The Trump administration has moved to advance AI development by reducing regulations and pushing back against state-level AI rules. A voluntary approach fits that posture. But it may not satisfy the coalition now pressing the president from the right.
Breitbart News reported this week that a coalition of more than 60 MAGA-aligned groups sent an open letter to President Trump calling for AI regulation. The effort was organized by Humans First, a conservative group whose stated mission holds that "technology should serve humans first... not replace them."
Signers included Steve Bannon, the "War Room" host, along with conservative activists Amy Kremer and Brendan Steinhauser. Bannon told Axios the letter marked a turning point in the right's engagement with AI policy.
"This letter takes us next level. The letter lays out [that] we must have mandatory testing and government approval."
That language, "mandatory testing and government approval", runs well beyond the voluntary framework the White House was preparing to sign. The gap between the two positions is significant. A voluntary system asks companies to cooperate. A mandatory regime tells them they must.
The coalition's letter arrived the same week the signing was scheduled. Whether the timing influenced the postponement is unknown, but the White House has offered no competing explanation. The administration now finds itself caught between its deregulatory instincts and a grassroots right that views unchecked AI development as a genuine threat, not just to jobs, but to the kind of ordered society conservatives have long defended. It is not unlike other moments when internal policy disagreements have forced the administration to recalibrate.
The White House had invited a wide range of tech CEOs to attend the signing. That guest list, still unnamed publicly, was clearly intended to project industry buy-in. A ceremony flanked by the heads of major AI firms would have signaled that the private sector welcomed the administration's approach.
But staging a signing event with the very executives a voluntary framework is supposed to govern raises an obvious question: Who is the framework for? If the companies helped shape it and showed up to celebrate it, skeptics on the right have reason to wonder whether the order would impose any real accountability at all.
New AI systems, including models like Anthropic's Mythos, are advancing rapidly. The pressure from Trump's base reflects a real anxiety that the government is falling behind the technology it is supposed to oversee. A voluntary 90-day review window may sound reasonable on paper. Whether it means anything in practice depends entirely on enforcement that, by definition, a voluntary framework lacks.
The administration has not been shy about making dramatic course corrections when circumstances shift. Postponing a signing ceremony is far less dramatic than reversing a foreign policy position. But the silence around this delay, no stated reason, no rescheduled date, suggests the White House recognizes it has a problem it has not yet solved.
Abrupt schedule changes are not new to this White House. The administration has canceled planned diplomatic moves on short notice when conditions on the ground shifted. In domestic policy, executive actions have faced legal headwinds as well, with courts stepping in to block or delay implementation on issues from immigration to regulatory authority.
AI policy, though, presents a different kind of challenge. There is no opposing nation to negotiate with and no court injunction to blame. The tension here is internal, between a president who wants to unleash American innovation and a base that wants guardrails before the innovation outruns the country's ability to manage it.
Both sides have a point. Overregulation could hand the AI race to foreign competitors. Underregulation could leave American workers, financial systems, and critical infrastructure exposed to risks that no one fully understands yet. The executive order was supposed to thread that needle. Its postponement suggests the White House isn't confident it does.
The broader context of executive action facing resistance, whether from federal courts blocking border policy or from allied political groups demanding stronger measures, is now a recurring feature of governance in the second Trump term. The AI order is the latest example, and perhaps the most consequential in the long run.
The White House has not said when, or whether, the executive order will be rescheduled. The voluntary framework could be signed as drafted, revised to incorporate tougher provisions, or shelved indefinitely while the administration works through its internal debate.
What the delay has already accomplished is this: it has made visible a fault line on the right that most of Washington has ignored. The assumption in the media and in Silicon Valley has been that conservatives uniformly favor deregulation of AI. More than 60 MAGA-aligned groups just said otherwise, loudly and in writing.
The tech CEOs who were ready to stand behind the president at a signing ceremony will have to wait. The conservative grassroots groups demanding mandatory testing and government approval will keep pushing. And the White House will have to decide whose vision of AI governance it actually shares.
When the people who put you in office start writing open letters, a voluntary framework and a photo op with Silicon Valley executives may not be enough to answer them.
