Trump Signs AI Executive Order Creating Treasury-Run 'AI Cybersecurity Clearinghouse' and 30-Day Frontier-Model Reviews

On June 2, 2026, President Trump signed an executive order titled 'Promoting Advanced Artificial Intelligence Innovation and Security,' directing national-security and civilian agencies to increase scrutiny of frontier AI models and to bolster federal cyber defenses against AI-enabled threats. A central provision tasks the Treasury Department—working with the NSA, CISA and other agencies and required to stand up within roughly one month—to establish an 'AI cybersecurity clearinghouse,' a voluntary collaboration with industry and critical-infrastructure operators to scan, discover and validate software vulnerabilities and coordinate patch distribution.

Within 30 days, CISA and White House officials must issue Binding Operational Directives to expedite cyber defense of civilian federal systems and expand AI-enabled defensive tools extended to state, local and critical-infrastructure operators. The order creates a voluntary framework under which developers may submit 'covered frontier models' for federal testing up to 30 days before public release—reduced from a 90-day window in an earlier draft the president pulled back on May 21 over competitiveness concerns—applying, per adviser David Sacks, only to models representing a 'meaningful step-change in cyber capabilities.' A classified benchmarking process to assess models' cyber capabilities is due by August 1, 2026, and the Attorney General is directed to prioritize prosecutions under the Computer Fraud and Abuse Act and fraud statutes for AI-enabled intrusions.

Reactions split: the Business Software Alliance praised the 'voluntary and phased approach,' while Finite State's Doc McConnell warned classified benchmarking could delay getting capable models 'into the hands of cyber defenders who can put them to use today.'