Patch window is officially dead as AI finds bugs faster than humans can squash them
AI technology is rapidly changing the landscape of vulnerability discovery in software security. Anthropic's Project Glasswing, utilizing the Claude Mythos AI model, has demonstrated the ability to find critical vulnerabilities faster than traditional methods can patch them. This shift raises significant concerns for security teams as the traditional patch window is effectively eliminated.
- ▪Anthropic's Project Glasswing includes major tech companies like Amazon, Apple, and Google.
- ▪The AI model Mythos has discovered vulnerabilities that manual auditing has missed, including a bug in OpenBSD that went undetected for 27 years.
- ▪More than 99% of the vulnerabilities found by Mythos have not yet been patched.
Opening excerpt (first ~120 words) tap to expand
Pro Patch window is officially dead as AI finds bugs faster than humans can squash them Opinion By Shane Fry published 21 May 2026 AI-driven vulnerability discovery outpaces traditional patching defenses When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works. (Image credit: Shutterstock) Copy link Facebook X Whatsapp Reddit Pinterest Flipboard Threads Email Share this article 0 Join the conversation Follow us Add us as a preferred source on Google Newsletter Subscribe to our newsletter Anthropic's Project Glasswing has changed the math on vulnerability discovery, and software teams need to sit with the implications.Project Glasswing is an industry coalition, including Amazon, Apple, Google, Microsoft, Cisco, and others, built around…
Excerpt limited to ~120 words for fair-use compliance. The full article is at TechRadar.
Discussion
0 commentsMore from TechRadar
