Trump Executive Order Shifts AI Policy Toward Active Cybersecurity Role
President Trump signed an executive order on June 2 that marks a departure from the administration's previous "light-touch" approach to AI regulation. The order, "Promoting Advanced Artificial Intelligence Innovation and Security," establishes a federal framework for reviewing frontier AI models before their public release and directs multiple agencies to upgrade the nation's cyber defenses.
The shift reflects growing concern about frontier models' ability to identify cyber vulnerabilities at unprecedented speed and precision-capabilities that bad actors could exploit.
The voluntary review framework
The order's centerpiece is a voluntary system allowing AI developers to submit frontier models for confidential cybersecurity review up to 30 days before public release. The Commerce Department's Center for AI Standards and Innovation will work with federal agencies to establish evaluation criteria and select "trusted partners" from the private sector to help identify vulnerabilities.
The framework explicitly prohibits mandatory licensing, government veto power over model releases, or control over launch timing. Developers retain final say on whether to release a model.
Key negotiation points for developers will include intellectual property protections, confidentiality terms, who can access their models during review, and the review period's length.
Federal cybersecurity upgrades
The order requires multiple agencies to act within 30 days. The Department of Homeland Security must release directives that prioritize cyber-readiness for federal systems, establish programs for AI-enabled cybersecurity tools, and provide access to frontier models for state and local governments and critical infrastructure operators like hospitals and banks.
The Treasury Department must establish an "AI cybersecurity clearinghouse" to coordinate vulnerability assessments across the government and private sector. The Office of Management and Budget will determine whether federal grant money can support AI vulnerability detection development.
Within 60 days, federal agencies must create a classified benchmarking process to assess frontier models' cyber capabilities, and the U.S. Space Force must prioritize hiring cybersecurity specialists.
Enforcement priorities
The order directs the attorney general to prioritize prosecution of crimes where perpetrators used AI, including identity fraud and computer fraud. This continues a federal acknowledgment that AI amplifies criminal capability.
What government workers should know
For federal employees managing AI policy or procurement, the order creates new obligations and opportunities. Agencies must prepare to evaluate frontier models against emerging benchmarks, which could become procurement standards for government contracts.
Standards developed through this process may influence broader industry guidance, including the National Institute of Standards and Technology's Cybersecurity Framework. Government workers should monitor how threshold definitions-particularly what qualifies as a "covered frontier model"-get established, as these will shape implementation across agencies.
State and local government officials, along with operators of critical infrastructure, will gain access to new federal cybersecurity tools and frontier models. Agencies should begin planning how to integrate these resources into existing security operations.
Learn more about AI for Government and explore resources on AI for cybersecurity professionals to understand how these policy changes affect your organization.
Your membership also unlocks:
