Trump’s New AI Order Seeks Voluntary Model Reviews and Faster Cyber Defenses

President Donald J. Trump signed an executive order on Tuesday that directs federal agencies to accelerate cybersecurity upgrades across government systems while inviting AI developers to share their most powerful models for review before wider release. The move, issued just hours ago, balances a push for rapid innovation with new attention to national security risks from advanced systems.

The document, titled “Promoting Advanced Artificial Intelligence Innovation and Security,” opens with a clear statement of intent. “Advanced AI capabilities make our Nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies,” it reads. Short. Direct. The order rejects heavy regulation. It credits previous steps to cut bureaucratic hurdles and instead calls for close work with private industry.

Agencies now face tight deadlines. Within 30 days the Committee on National Security Systems and the Secretary of War must prioritize defenses for their respective networks. The Secretary of Homeland Security, acting through the Cybersecurity and Infrastructure Security Agency, must issue binding operational directives. These steps target civilian federal systems first. They also expand access to AI-powered defensive tools for state and local governments, rural hospitals, community banks and local utilities. The goal is protection of vital functions against external threats.

One concrete creation stands out. The Secretary of the Treasury, working with the National Cyber Director, NSA and CISA, will form an AI cybersecurity clearinghouse. This body will coordinate vulnerability scanning, discovery, validation and patch distribution. Participation remains voluntary for industry and critical infrastructure operators. Funding from existing federal grant programs could flow toward developers focused on vulnerability detection. Hiring pathways for cybersecurity specialists will expand through the Office of Personnel Management.

The order’s most discussed element involves frontier models. Within 60 days officials must develop a classified benchmarking process. The Director of the National Security Agency, in consultation with others, will set the threshold for what counts as a “covered frontier model.” Developers can then opt into a voluntary framework. They may submit models for government assessment up to 30 days before sharing them with other trusted partners. Access comes with strict confidentiality, cybersecurity and intellectual-property safeguards. The process aims to select additional trusted partners and strengthen defenses for critical infrastructure.

Yet the language draws a firm line. “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 administration wants collaboration. It does not want to slow American labs in their race against global competitors.

The Justice Department receives marching orders too. Attorney General must prioritize enforcement of computer fraud and abuse statutes against anyone who uses AI to gain unauthorized access, damage systems or commit other crimes. That includes AI agents deployed for unlawful data extraction.

Industry groups responded quickly. The Software & Information Industry Association welcomed the focus on both innovation and security. “Today’s Executive Order recognizes that innovation and security go hand in hand,” said Paul Lekas, its executive vice president for global public policy, in a statement. Business Roundtable CEO Joshua Bolten echoed the sentiment. His organization called the order “an important step forward in protecting critical infrastructure while supporting the safe deployment of advanced models and preserving the ability to innovate,” according to its release.

Analysts note the order evolved from an earlier draft. A version considered in May reportedly included a longer 90-day review window. Concerns that such a delay would harm U.S. competitiveness against China led to its postponement. The final 30-day period reflects that adjustment. Council on Foreign Relations senior fellow for artificial intelligence described the document as a shift toward federal oversight while maintaining a deregulatory stance.

Implementation will test coordination. The order tasks the Treasury, Defense, Homeland Security, Commerce and White House offices with joint work. It relies on existing authorities and available appropriations. No new rights or benefits are created for outside parties. Costs for publication fall to the Department of War.

Observers on X reacted within minutes of the announcement. Some praised the emphasis on voluntary engagement and cyber hardening. Others worried the scope remains too narrow. One policy expert argued the framework should address risks beyond cyber capabilities, including internal model development and potential model theft by adversaries. “The government deserves to know what capabilities AIs have at the same time companies know, not just 30 days before commercial deployment,” the post noted.

The timing matters. AI models continue to gain sophisticated offensive and defensive abilities. Recent demonstrations, including systems that exhibit superhuman performance in narrow domains, have heightened concerns among national security officials. The order explicitly ties American leadership in AI to both economic strength and military advantage. It positions collaboration with industry as the path to deploy the “best and most secure technology” against threats.

Details on the classified benchmarking process will emerge slowly. So will the exact criteria for covered frontier models. For now the order sets a direction. Agencies must move fast. Industry has an invitation to participate on its own terms. The bet is that voluntary mechanisms, combined with upgraded federal defenses and targeted enforcement, can address emerging risks without choking the innovation that has kept the United States ahead.

Whether that balance holds will become clearer in the coming months as directives roll out and companies decide whether to engage with the new clearinghouse and review framework. The executive order offers a starting framework. Its success depends on execution across multiple departments and genuine partnership with the private sector that develops these systems.