Trump Signs AI Executive Order Requiring Early Government Access to Frontier Models

President Donald Trump on Tuesday signed an executive order directing artificial intelligence companies to provide the federal government with early access to their most advanced models ahead of public release, framing the directive around national security and the assessment of AI-related cyber capabilities.

The order asks companies, on a voluntary basis, to participate in a government-led benchmarking process designed to evaluate a model's "advanced cyber capabilities" and determine whether it qualifies as a "covered frontier model." Once a model clears that threshold, companies would be expected to give the government access up to 30 days before a broader commercial release.

The order also enables the government to help designate the "trusted partners" that receive that early access, though it stops short of creating a mandatory licensing or preclearance framework.

"Nothing in this section shall be construed to authorize the creation of a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models, including frontier models," the order states.

Trump signed the order in private. The signing follows a postponed ceremony with prominent tech executives, which Trump called off weeks earlier after telling reporters he "didn't like certain aspects of it."

The order arrives at a consequential moment in the AI industry. Anthropic, the developer of the Claude family of models, said Monday that it had confidentially filed with the Securities and Exchange Commission for an initial public offering. Rival OpenAI is also preparing for a potential offering this year, according to reports. Elon Musk's SpaceX, which owns his AI lab SpaceXAI, is reportedly poised to reach public markets as soon as next week at a valuation that could exceed $1 trillion.

The directive also carries specific implications for the Department of Defense, which the order calls on to prioritize the cyber defense of its information systems. The DOD has taken a complicated posture toward Anthropic in particular, having previously labeled the startup a supply chain risk — a designation that prohibits defense contractors from using the company's technology. Anthropic sued the Trump administration to reverse that classification, and that litigation remains ongoing.

Anthropic's prominence in the policy conversation stems in part from its Claude Mythos Preview model, which the company announced earlier this year and which drew attention for its ability to identify weaknesses and security flaws in software. Anthropic limited its initial rollout through a cybersecurity initiative called Project Glasswing, which it said Tuesday it was expanding. The model's launch prompted meetings between Anthropic and senior administration officials, including White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent.

The order also outlines several timeframes for agencies to develop additional guidance, though it is thin on specifics.

The tech industry's influence over the administration's AI posture has been on display throughout the process. Venture capitalist David Sacks, a longtime ally of Musk's, served as the White House's first crypto and AI czar before that role ended earlier this year. Sacks, along with Musk and Meta CEO Mark Zuckerberg, reportedly lobbied the Trump administration last month against the earlier version of the AI executive order.

Whether the voluntary participation framework will draw broad industry cooperation — and how the government intends to enforce access timelines without a mandatory mechanism — will likely define the order's practical impact in the months ahead.

Disclaimer