Trump Signs Executive Order Granting Government Early Access to Advanced AI Models

Washington - June 2, 2026

President Donald Trump signed an executive order Tuesday directing federal agencies to establish a voluntary framework with AI developers, granting the government access to the most advanced AI systems up to 30 days before their release to other partners - a move driven by mounting national security concerns over the cybersecurity capabilities of frontier AI models.

The order, signed at the White House, tasks the Treasury Department, the National Security Agency, and the Cybersecurity and Infrastructure Security Agency (CISA) with developing a classified benchmarking process to assess which AI models qualify as "covered frontier models" - a new designation for systems deemed to carry advanced cyber capabilities that warrant government review.

Voluntary, Not Mandatory

The executive order draws a firm legal line: nothing within it authorizes a mandatory licensing, preclearance, or permitting requirement for AI models. Participation by companies such as Anthropic, OpenAI, and Google remains strictly voluntary - a deliberate concession to an industry that pushed back hard against an earlier draft calling for a 90-day review window. The final order cuts that period to 30 days.

The Mythos Effect

The order arrives in the wake of Anthropic's Claude Mythos Preview - a model that demonstrated unprecedented capability in identifying software vulnerabilities - which triggered a series of high-level meetings between Anthropic executives and senior Trump administration officials, including White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent. The Pentagon, notably, had previously labeled Anthropic a supply chain risk shortly before Mythos launched.

A Three-Track Response

Beyond the frontier model framework, the order launches three parallel tracks:

First, a cybersecurity clearinghouse under the Treasury Department - built in voluntary collaboration with the AI industry and critical infrastructure operators - to coordinate vulnerability scanning, discovery, and patch distribution across sectors including rural hospitals, community banks, and local utilities.

Second, a criminal enforcement directive instructing the Attorney General to prioritize prosecution under existing federal statutes against any individual who uses AI to illegally access or damage computer systems.

Third, a federal hiring expansion through the Office of Personnel Management, widening pathways for cybersecurity specialists within the US Tech Force within 60 days.

Strategic Context

The order reflects a broader balancing act at the heart of Trump's AI policy: accelerating American AI dominance while confronting the security risks that same dominance creates. The administration has consistently resisted the regulatory approach of its predecessor, instead favoring industry collaboration - a posture this order formalizes for the first time in the context of frontier AI security.

EcoPulse24 Analysis

When AI Becomes a Weapon, Governments Move

What Trump signed today is not routine regulatory housekeeping - it is an official acknowledgment that artificial intelligence has crossed a threshold. It is no longer merely a productivity tool; it has become a strategic variable in the national security equation. The same administration that dismantled Biden-era AI constraints on day one, under the banner of "innovation first," now finds itself stepping back - not to restrict the industry, but to absorb its risks before those risks absorb it.

Anthropic's Mythos was not simply another model release - it was a moment of revelation. When the model demonstrated an unprecedented ability to detect and exploit software vulnerabilities, the debate shifted overnight from research lab corridors to Pentagon briefing rooms and Wall Street risk desks simultaneously. Notably, the Department of War had already labeled Anthropic a "supply chain risk" before the model launched - a remarkable designation for an American company by its own government, exposing the depth of institutional alarm that Mythos triggered within the national security establishment.

The voluntary framework at the heart of this executive order carries a structural gamble. The government is asking AI companies to hand over access to their most sensitive models without any legal mechanism to compel compliance. The order's effectiveness therefore hinges entirely on whether the industry trusts that participation will not be weaponized as justification for future mandates. That trust is difficult to guarantee in a politically volatile environment - and AI executives know it.

There is an economic dimension that most coverage is missing: the establishment of an AI cybersecurity clearinghouse under Treasury is, in substance, the seed of a government-private hybrid system for managing AI risk at scale. If it works, it becomes a exportable model - either as an international standard or as leverage in global AI governance negotiations. Washington understands that whoever sets the security benchmarks today controls the rules of the game tomorrow.