SANS Internet Stormcenter Daily Cyber Security Podcast (Stormcast) SANS Stormcast Monday, February 9th, 2026: Azure Vulnerabilties; AI Vulnerability Discovery; GitLab AI Gateway Vuln
6 snips
Feb 9, 2026 Coverage of four patched Azure vulnerabilities affecting services like Front Door and Functions. Discussion of AI tools finding zero-days and the debate over their usefulness. Review of Anthropic’s study claiming hundreds of LLM-discovered high-impact flaws. Report on a GitLab AI Gateway flaw that allowed authenticated code execution on on-prem installs.
AI Snips
Chapters
Transcript
Episode notes
Critical Azure Vulnerabilities Patched
- Microsoft patched four Azure vulnerabilities, three rated critical, affecting Front Door, Functions, and Arc.
- Johannes Ulrich notes customers generally need not act because Microsoft already mitigated these issues.
AI Finds Missed Open-Source Flaws
- Anthropic used their Opus 4.6 model to discover about 500 high-impact issues in open-source projects.
- They emphasize skilled operators and human validation to reduce low-quality AI reports and false positives.
AI Complements Traditional Testing
- Anthropic highlights GoScript and OpenSC as fruitful targets because past fuzzing still missed issues.
- This suggests AI can complement but not replace traditional security testing approaches.