Trump has signed an executive order pushing U.S. security agencies to move faster on advanced artificial intelligence, placing AI at the center of Washington’s efforts to defend national security, federal and critical infrastructure systems from emerging cyber threats.
The order, titled “Promoting Advanced Artificial Intelligence Innovation and Security,” directs federal agencies to harden sensitive networks, expand AI-enabled cyber defenses and work more closely with private AI companies as the technology becomes increasingly important to national security.
Pentagon ordered to speed defenses for AI-era threats
The order gives the Department of War 30 days to prioritize the cyber defense of its information systems, while the Committee on National Security Systems must take similar action for national security networks.
The Department of Homeland Security, through the Cybersecurity and Infrastructure Security Agency, is also directed to issue guidance to speed up defenses for civilian federal systems.
The order says those efforts should include wider use of AI-enabled security tools and expanded access to cybersecurity services for federal agencies, state and local authorities, and operators of critical infrastructure, including rural hospitals, community banks and local utilities.
The move reflects the administration’s view that advanced AI can strengthen U.S. defenses while also creating new national security risks that require faster coordination across federal agencies.
Washington seeks early look at powerful AI models
A major part of the order focuses on powerful frontier AI models that could carry advanced cyber capabilities.
The Treasury Department, the Department of War through the National Security Agency, and Homeland Security through CISA are directed to create a classified benchmarking process to assess those risks. The process will help determine when an AI system should be treated as a “covered frontier model” under the order.
The White House is also calling for a voluntary framework with AI developers that would allow companies to work with the federal government before releasing highly capable models to trusted partners.
Under that framework, developers could provide government access to covered models for up to 30 days, subject to confidentiality, cybersecurity and intellectual-property protections.
AI push comes without mandatory model approval
The order stops short of creating mandatory licensing, preclearance or permitting rules for new AI models, a point likely aimed at reassuring the technology industry that Washington is not seeking to slow development through formal approval requirements.
At the same time, it directs Treasury, NSA, Homeland Security and CISA to form an AI cybersecurity clearinghouse with industry and critical infrastructure operators to coordinate vulnerability scanning, validation and patch distribution.
The Attorney General is separately ordered to prioritize cases involving people who use AI to illegally access, damage or breach computer systems.
The order marks a more aggressive federal push to fold advanced AI into U.S. national security operations, while drawing early guardrails around the most powerful systems before they are widely deployed.
