White House Launches 'Gold Eagle,' a Federal Clearinghouse to Patch the Flood of AI-Discovered Software Flaws

The White House on Tuesday launched a new federal clearinghouse, dubbed "Gold Eagle," designed to gather software vulnerability findings from across government and industry and coordinate fixes at what officials called a speed and scale never before attempted. The effort is a direct response to a surge in security flaws being uncovered by artificial intelligence models — a wave that is straining the ability of human security teams to keep up.

According to the administration, Gold Eagle consolidates vulnerability discoveries, prioritizes the most critical flaws and coordinates remediation across the nation's critical infrastructure. The White House said the clearinghouse has already begun receiving intelligence on vulnerabilities and prioritizing patches. Officials framed the launch in stark terms, describing it as an effort to put America's cyber defenses on what one description called a "wartime footing" against a rising tide of automated threats.

The initiative brings together several federal agencies, including the Cybersecurity and Infrastructure Security Agency, or CISA, the Department of the Treasury and the Department of Defense, working alongside open-source software organizations and critical infrastructure operators. Gold Eagle was established under a June 2 executive order titled "Promoting Advanced Artificial Intelligence Innovation and Security," designated EO 14409, which directed the federal government to promote the secure use of advanced AI systems while accelerating American innovation in the field.

The clearinghouse is not a regulatory hammer. Officials stressed that Gold Eagle does not compel companies to fix flaws directly; instead, it functions as a coordination mechanism that routes intelligence to the partners best positioned to act. It builds on existing CISA programs — including the agency's coordinated vulnerability disclosure process, its Known Exploited Vulnerabilities catalog, the industry-standard CVE system and the National Institute of Standards and Technology's National Vulnerability Database — layering faster detection and prioritized response on top of that infrastructure. AI developers, with Anthropic identified as a likely participant, are expected to feed findings into the system.

The launch comes amid mounting evidence that AI tools have become potent bug-hunters, capable of scanning vast codebases and surfacing flaws that once took human researchers months to find. That capability is double-edged: the same models that help defenders can help attackers, and the sheer volume of newly discovered vulnerabilities has overwhelmed traditional patching pipelines. Supporters argue a central clearinghouse is exactly what is needed to triage that flood; skeptics will watch closely to see whether Gold Eagle can move fast enough — and whether sensitive vulnerability information can be shared without itself becoming a target. For now, the administration is betting that coordination, not new mandates, is the fastest path to closing the gaps.