Anthropic’s Claude Opus 4.6 Raises the Bar for AI Cybersecurity: 500+ High-Severity Open-Source Vulnerabilities Discovered
What happens when an AI stops brute-forcing inputs and starts reading code like a senior engineer with a photographic memory? In early February 2026, Anthropic answered that question with Claude Opus 4.6—a model that didn’t just “find bugs,” it reasoned its way to over 500 previously unknown, high-severity vulnerabilities in widely used open-source projects. In…