Anthropic expands Project Glasswing to 150 organizations in more than 15 countries

Anthropic is expanding Project Glasswing to roughly 150 organizations across more than 15 countries, using the Claude Mythos Preview model to automatically scan software for vulnerabilities; the initiative has identified over 10,000 high- and critical-severity flaws. Anthropic will make tools available to trusted security teams and warns this acceleration in vulnerability discovery will strain patch pipelines and increase systemic risk if advanced AI cyber-capabilities are released without robust safeguards.