Tagged “AI”
AI teams inside ordinary companies now run HPC-class machines holding models and training data. Why the enterprise security stack misses them, and how the NIST SP 800-223 zone architecture applies.
How to threat model AI systems: STRIDE extended for LLM applications, the OWASP LLM Top 10 mapped to design-time decisions, and a worked RAG example.
A walkthrough of one representative afternoon: a whiteboard photo becomes an editable architecture diagram, a written design explanation, a STRIDE threat model with mapped controls, and a signed design record before the end of the day.
The workshop is the scaling bottleneck of threat modeling, not the analysis. How an async-first process covers more systems with less calendar: structured intake, diagrams from what exists, proposed threats, and one short conversation where judgment actually matters.
Most agent security guidance assumes a foundation you may never have finished building. This is the tier nobody wrote down.
AI now finds vulnerabilities faster than teams can verify, disclose, and patch them, and attackers hold the same tools. The monthly patch cycle has quietly stopped being enough. Here is what belongs on a 2026 vulnerability management roadmap.