Anthropic Launches Project Glasswing for AI-Driven Vulnerability Detection

Anthropic has launched Project Glasswing, an initiative utilizing its Claude Mythos Preview AI to autonomously detect and remediate previously undiscovered vulnerabilities in critical software. The project aims to enhance security by leveraging AI-driven analysis without specifying targeted software types, affected versions, or identified CVEs. No technical implementation details, such as scanning methodologies or fix deployment mechanisms, were disclosed in the announcement. The initiative was announced without a precise launch date or geographic focus. The reported impact centers on improving vulnerability discovery and patching efficiency in software systems.