White House creates AI security group to protect critical infrastructure

The White House created GOLD EAGLE on Tuesday, a Treasury-led clearinghouse built with the Department of Homeland Security’s Cybersecurity and Infrastructure Security Agency and the Department of War to speed sharing of cybersecurity vulnerabilities uncovered by advanced AI systems. The goal is to reduce duplicate scanning, line up remediation faster and push actionable threat information to operators before weaknesses in widely used AI tools can be exploited.

The White House designed the system to help critical-infrastructure operators, federal agencies and state and local governments exchange information on flaws found in their own tools or networks, with special attention to rural hospitals, community banks and local utilities that often have fewer in-house cyber defenses.

The initiative follows Executive Order 14409, signed June 2, 2026, which ordered agencies within 30 days to prioritize cyber defense for National Security Systems, Department of War information systems and civilian federal systems. The order also directed DHS and CISA to issue binding operational directives and other guidance, and put the AI cybersecurity clearinghouse in voluntary coordination with industry and critical-infrastructure operators to identify and remediate software vulnerabilities at scale. It explicitly did not authorize mandatory licensing, pre-clearance or permitting requirements for AI models.

Sean Cairncross, the White House cyber director, included developers of open-source models in the effort, though the administration did not name them individually. White House materials extended AI-enabled cybersecurity tools and services to federal agencies, state and local authorities and critical-infrastructure operators, while establishing a classified benchmarking process for covered frontier models and a voluntary framework that gives trusted partners secure early access.

Trump’s June 5 National Security Presidential Memorandum on AI in the national security enterprise said AI can help protect warfighters, minimize civilian harm and maintain technical overmatch against adversaries. In April 2026, OpenAI was scaling its Trusted Access for Cyber program to thousands of verified defenders and hundreds of teams defending critical software, while Anthropic announced Project Glasswing on April 7, 2026, with major technology and finance companies to secure critical software. Anthropic says AI models can now find high-severity vulnerabilities at scale, leaving verification, disclosure and patching as the harder task.