Anthropic Expands Project Glasswing to Secure Critical Infrastructure

Anthropic announced on June 24 that it is expanding Project Glasswing, its initiative to secure critical software using AI, by onboarding 150 additional critical infrastructure organizations. Project Glasswing, launched in 2025, gives participating organizations access to Anthropic's Mythos cybersecurity AI to identify and remediate vulnerabilities in their systems.

Mythos has already identified over 10,000 high-severity vulnerabilities across every major operating system and web browser according to Anthropic's internal testing. The AI model has discovered security flaws as old as 27 years in systems maintained by small volunteer teams, demonstrating its ability to find issues that traditional security tools might miss.

The expansion includes a notable cross-section of participants: five Fortune 100 companies, three state-level governments, and several critical infrastructure providers including energy, transportation, and healthcare organizations. Anthropic states that organizations participating in Project Glasswing have reported a 40% reduction in unpatched critical vulnerabilities within their systems.

The strategic significance of this expansion is profound. As AI systems become increasingly capable, they're becoming both targets and tools in cybersecurity. Mythos represents a defensive application of AI — using advanced models to find and fix security flaws rather than exploit them.

For organizations in the MENA region undergoing rapid digital transformation, Project Glasswing offers a model for proactive cybersecurity: combining AI-powered vulnerability discovery with expert security guidance to address the unique challenges of operating in complex geopolitical environments.