#OSS-Security

Anthropic Project Glasswing: AI Vulnerability Discovery Challenges

Anthropic's Project Glasswing identifies 23,000 potential vulnerabilities via the Mythos AI model, highlighting the massive gap between discovery and patching.