President Donald Trump signed an executive order on Tuesday, June 2, 2026 [1], requiring AI companies to grant the federal government early access to advanced models.
This move establishes a direct oversight mechanism for the U.S. government to evaluate the security risks of emerging technologies before they reach the public. By securing early access, the administration aims to prevent national security vulnerabilities and ensure the state can counter threats using the most current tools.
The order directs the Attorney General to prioritize the prosecution of AI-related crimes [2]. This legal focus accompanies the administration's effort to maintain a competitive edge in technology while managing the inherent risks of rapid deployment.
According to the order, the goal is to "ensure the best and most secure technology is deployed rapidly to confront any and all threats to our country" [3]. The administration said that while AI can strengthen the U.S., it also creates significant risks without government oversight [4].
The mandate specifically targets advanced models that could pose systemic security risks [2]. The administration said that the order is designed to protect critical infrastructure, including local banks and hospitals, from AI-driven threats [5].
Reports on the status of the order have been inconsistent. While some sources confirm the signing on June 2 [1], other reports suggested the order was postponed or called off due to concerns that such regulations might inhibit economic growth [6, 7]. However, the primary directive focuses on the necessity of government oversight to mitigate national security risks [4].
“"AI would make the U.S. stronger but it also poses a national security risk without government oversight"”
This executive order signals a shift toward a 'security-first' approach to artificial intelligence, prioritizing state oversight over the unrestricted speed of private-sector development. By requiring early access to models, the U.S. government is attempting to treat advanced AI as a dual-use technology—similar to nuclear or biological agents—where the potential for catastrophic misuse necessitates government vetting before widespread release.
