AI discovered a 14-year-old vulnerability that's been hiding in plain sight across internet infrastructure. The flaw potentially affected millions of systems, sitting dormant in widely-deployed code for over a decade. This raises questions about how many other ancient bugs are lurking in foundational software that human audits missed. The real story here isn't just the discovery, but what it means for automated security analysis at scale. Traditional code review clearly failed to catch this for 14 years. Now we're seeing AI models trained on vulnerability patterns actually finding needles in massive legacy codebases. The technical implications are huge - if AI can spot decade-old bugs, we might need to systematically re-audit critical infrastructure with these tools. Also makes you wonder what the actual attack surface was and whether it was already being exploited in the wild before detection.