Anthropic's AI, Claude Mythos, uncovered over 10,000 critical software vulnerabilities within a month, showcasing its unprecedented potential for cybersecurity advancements.
Mythos was tested on systemically important software
May 23, 2026
01:19 pm
What's the storyAnthropic, the AI start-up that developed Claude Mythos, has revealed that its advanced artificial intelligence (AI) system has detected over 10,000 critical cybersecurity flaws in just a month.
The revelation was made during Project Glasswing, an initiative where select companies like Amazon and Google were given access to Mythos.
Anthropic uncovered 10,000 vulnerabilities through Project Glasswing, driving urgent patching efforts and stronger cyber defenses.
Anthropic's AI model Claude Mythos Preview, working with about 50 partners as part of Project Glasswing, has found over 10,000…
Anthropic's Project Glasswing used its Claude Mythos Preview AI to find over 10,000 critical software vulnerabilities, including…
Project Glasswing partners used Claude Mythos to find 1,094 confirmed high-severity flaws across 1,000+ open-source projects.…
Last week, Anthropic pulled back the curtain on Claude Mythos Preview, an AI model so capable at finding and exploiting software…
The AI that's good at finding security flaws is doing its thing in the toughest arena out there.
