We finally have a look at the actual draft executive order on AI safeguards that President Donald Trump suddenly decided not to sign on Thursday.
According to a draft of the order sent to Gizmodo from a source with knowledge of the negotiations, the executive order would have created a framework for AI companies to voluntarily provide access to frontier AI models to the federal government up to 90 days before wide release to “strengthen the cybersecurity of critical infrastructure.” Apparently it was leaked far and wide, with Politico being the first to publish it.
Leaks in recent days suggested that the entire framework will be largely voluntary, but the actual draft places considerable emphasis on that point.
“Nothing in this section shall be construed to authorize the creation of a mandatory government licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models, including frontier models,” the draft reads.
So, it’s a bit puzzling why the Trump administration would change its mind about the voluntary program at the last minute.
Some reports claimed that this happened because the top AI CEO was unable to reach the White House for the autograph and photo session. Instead, he may have sent lower-level officials, according to Politico.
The outlet also reports that the administration’s AI czar, David Sachs, raised the industry’s concerns directly with Trump on Wednesday night. Sachs reportedly made the case that AI companies are already cooperating with the government and that any federal review process could slow innovation and give China the edge in an AI arms race. He also expressed fear that any voluntary framework would someday turn into mandatory government review.
For his part, Trump appeared to echo that argument.
“I didn’t like some aspects of it,” Trump told reporters Thursday morning. “I think it gets in the way … we’re leading China. We’re leading everybody, and I don’t want to do anything that gets in the way of that.”
The White House did not immediately respond to a request for comment.
Still, this is the latest example of the Trump administration essentially giving the AI industry exactly what it wants. So far, the Trump administration has taken a pragmatic approach to regulating AI. AI and billionaire ally Trump signed an executive order in December aimed at curbing what his administration describes as overly burdensome state regulations in the name of national and economic security.
The Trump administration took no significant action against an AI company until earlier this year, when the startup designated Anthropic as a supply chain risk after the Pentagon refused to allow it to use its technology for “any lawful purpose.” Anthropic wanted a settlement to prevent the government from using its products in autonomous weapons or for domestic surveillance. In the Pentagon’s view, this was too much to ask.
Amidst the federal oversight void, AI companies like OpenAI and Anthropic are trying to get ahead of things by lobbying for their preferred state laws.
But not all conservatives are entirely satisfied with how Trump is handling AI.
Longtime Trump ally Steve Bannon and more than 60 other conservative leaders recently sent a letter to the president urging him to sign the executive order, while also demanding greater government oversight of “potentially dangerous” frontier AI models.
Here is the entire draft of the executive order:
executive Order
– – – – – – –
Promoting advanced artificial intelligence innovation and security
In exercise of the authority vested in me as President by the Constitution and laws of the United States, it is ordered:
seconds. 1. Objective. The United States continues to lead the world in Artificial Intelligence (AI) because of the immense talent and innovation of our AI industry, and because we refuse to stifle this innovation with overly burdensome regulation. My Administration has brought tremendous technological development and economic investment in AI by reducing the bureaucratic barriers imposed by the prior administration on America’s AI developers and researchers and instead encouraging AI innovation and accelerating responsible AI adoption in government and industry.
Advanced AI capabilities make our nation stronger, but also present new national security considerations that require coordinated action across executive departments, agencies, and constituents. As these capabilities develop, my Administration will continue to work closely with industry to ensure that the best and safest technologies are rapidly deployed to confront any and all threats to our nation. We will continue to lead an America First cybersecurity effort that enhances both our national security and our global AI dominance.
It is the policy of the United States to promote AI innovation and security by working closely with the private sector to modernize government and private sector information systems and harden them against external threats; protecting American talent and intellectual property from exploitation and theft by adversaries; and developing America’s advanced AI-enabled capabilities.
seconds. 2. Upgrading US systems for advanced AI.
(a) Within 30 days after the date of this order, the Committee on National Security Systems shall, by taking appropriate and expeditious action consistent with the purpose of this order, prioritize the cyber defense of national security systems, as defined in 44 U.S.C. 3552(b)(6)(A).
(b) Within 30 days after the date of this order, the Secretary of War will prioritize the cyber defense of War Department information systems by taking appropriate and prompt action consistent with the purposes of this order.
(c) Within 30 days after the date of this order, the Secretary of Homeland Security, through the Director of the Cybersecurity and Infrastructure Security Agency (CISA), in consultation with the Director of the Office of Management and Budget (OMB), the Assistant to the President for National Security Affairs, and the National Cyber Director, shall issue binding operational directives and other guidance as appropriate:
(i) Accelerating and prioritizing the cybersecurity of civilian Federal Government information systems to protect our Nation’s vital functions;
(ii) establish or expand federal programs and cybersecurity services that enhance AI-enabled defensive tools; And
(iii) Where appropriate, facilitate access to cybersecurity tools and services, including covered frontier models, for Federal agencies, State and local authorities, and critical infrastructure operators such as rural hospitals, community banks, and local utilities.
(d) Within 30 days after the date of this order, the Secretary of the Treasury, in consultation with the National Cyber Director, the Secretary of War, through the Director of the National Security Agency (NSA), and the Secretary of Homeland Security, through the Director of CISA, shall create an AI Cybersecurity Clearinghouse, in voluntary collaboration with the AI industry and critical infrastructure operators, to coordinate scanning for software vulnerabilities and deconflict such vulnerabilities. Verifies and coordinates and prioritizes remediation and distribution of vulnerability patches.
(e) Within 30 days after the date of this order, the Director of OMB, in coordination with the National Cyber Director and the Director of CISA, will determine whether any Federal grant programs have available and relevant funding that can be directed to applicants developing advanced AI vulnerability identification.
(f) Within 60 days of the date of this order, the Director of the Office of Personnel Management will expand U.S. Tech Force information cybersecurity specialist recruitment and placement pathways.
seconds. 3. Secure Frontier Model deployment. Within 60 days after the date of this order, the Secretary of the Treasury, the Secretary of War, through the Director of NSA, and the Secretary of Homeland Security, through the Director of CISA, in consultation with the White House Chief of Staff, through the National Cyber Director, the Assistant to the President for Science and Technology (APST), and the Director of the National Institute of Standards and Technology, and in coordination with other executive departments and agencies, as appropriate, will:
(a) Develop and maintain a classified benchmarking process to assess the advanced cyber capabilities of AI models and determine the threshold at which AI models should be designated “covered frontier models” for purposes of this order, sharing such assessments with AI developers and researchers as appropriate. Such determination will be made by the Director of NSA, in consultation with the National Cyber Director, APST, the Director of CISA, and other War Department representatives, as appropriate.
(b) Create a voluntary framework with AI developers through which developers will be able to:
(i) to engage the U.S. Government in determining whether models under development meet the designation of “covered frontier model”;
(ii) plan to release such models to other trusted partners for a period of 90 days before providing the Federal Government with access to the covered frontier models subject to appropriate privacy, cybersecurity, insider risk, and intellectual property protection, use, and nondisclosure requirements; And
(ii) Collaborate with the federal government to select trusted partners who will have early access to covered frontier models to foster secure innovation and strengthen the cybersecurity of critical infrastructure.
(c) Nothing in this section shall be construed to authorize the creation of a mandatory government licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models, including frontier models.
seconds. 4. protection against criminal actors. The Attorney General will prioritize enforcement of 18 USC 1028, 18 USC 1030, 18 USC 1343, and all other applicable federal criminal laws against anyone who uses an AI to illegally access or cause harm to a computer without authorization, or who uses an AI engaged in such illegal access in furtherance of any other crime. This includes breaching any public or private information technology system, or employing AI agents to unlawfully access data or information, which is then used for a criminal or unlawful purpose.
seconds. 5. general Provisions. (a) Nothing in this order shall be deemed to impair or otherwise affect:
(i) any authority conferred by law on any executive department or agency or its head; Or
(ii) Functions of the Director of OMB relating to budgetary, administrative, or legislative proposals.
(b) This order shall be implemented consistent with applicable law and subject to the availability of appropriations.
(c) This order is not intended to create any right or benefit, substantive or procedural, enforceable at law or in equity by any party against the United States, its departments, agencies, or entities, its officers, employees, or agents, or any other person.
(d) The cost of publication of this order shall be borne by the War Department.
the White House
XX May, 2026.
<a href