White House Launches Cybersecurity AI Initiative for Critical Infrastructure

Defense Secretary Pete Hegseth said Washington is taking a ‘wartime footing’ in cyberspace by using AI to detect and fix security flaws faster.

President Donald Trump holds up his signed AI initiative during a signing ceremony in the Oval Office of the White House on Dec. 11, 2025. Alex Wroblewski/AFP via Getty Images

The White House has launched an artificial intelligence (AI) and cybersecurity initiative to help identify software vulnerabilities and coordinate responses across critical infrastructure sectors, including finance, healthcare, and energy.

The initiative, called GOLD EAGLE, creates a clearinghouse for coordination between federal agencies and open-source software partners, allowing for faster patching of cybersecurity weaknesses with the assistance of AI.

“We are bringing a wartime footing to the cyber domain to relentlessly patch vulnerabilities,” Secretary of War Pete Hegseth said in the July 14 release, adding that the initiative is “leveraging frontier AI alongside top American innovators to safeguard our critical infrastructure and protect the homeland.”

The White House, along with at least six agencies, will work with the private sector, the White House said.

U.S. officials increasingly view AI as a double-edged sword. Advanced AI systems developed by companies such as Anthropic and OpenAI can quickly identify software weaknesses, helping organizations fix them before they are exploited.

But they have also warned that criminals and foreign adversaries could use the same technology to find and exploit vulnerabilities in computer systems used by banks, hospitals, energy companies, and other critical infrastructure.

Pages from the Anthropic website and the company's logo are displayed on a computer screen in New York City on Feb. 26, 2026. Patrick Sison/AP

During a Senate Intelligence Committee hearing on June 11, Vice Chairman Mark Warner (D-Va.) said the National Security Agency had informed lawmakers that Anthropic’s advanced AI model, Mythos, was able to identify vulnerabilities across nearly all classified U.S. computer systems within hours rather than weeks.

Anthropic and the National Security Agency did not publicly confirm Warner’s account.

House Intelligence Committee Ranking Member Jim Himes (D-Conn.) later told The Epoch Times that he had been briefed on the model’s capabilities and described them as “very, very serious.”

Days later, Anthropic announced that it had received a U.S. government export-control directive requiring it to suspend access to its Fable 5 and Mythos 5 AI models for foreign nationals.

The company subsequently disabled access to both models for all users while questioning the government’s basis for the directive.

Cyber Defense

Officials said the program is already gathering reports of software security flaws, verifying whether they are real, and helping coordinate fixes across government and industry.

National Cyber Director Sean Cairncross said that leading AI developers and operators of critical infrastructure would participate in the program but did not identify all of the organizations involved.

Secretary of War Pete Hegseth delivers remarks before a NATO Defense Ministers meeting at the NATO headquarters in Brussels on June 18, 2026. John Thys/AFP via Getty Images

The new program follows the president’s executive order on AI, signed on June 2, which directed several federal agencies to work with AI companies and operators of critical infrastructure to improve cybersecurity.

The order also called for wider use of AI tools to improve cybersecurity while avoiding new licensing requirements for AI developers.

Ryan Morgan and Nathan Worcester contributed to this report.