38,000 AI-found vulnerabilities, 217 fixed: auditing Glasswing, Anthropic's bug-discovery machine

Seven months ago Anthropic claimed thousands of high-severity AI-discovered bugs. The independent audit is in: 37,966 findings, 7,349 public, 217 fixed, and an October scanner that emails maintainers fully model-generated reports with no human review. curl's maintainer got 45 reports for 6 low-severity confirmations. I pulled the tracker's live numbers myself: the month-by-month curve, top products, and the check for your own stack.