AI Security
Lessons from the field. Always landing on my feet.
Agent 365 Launch Playbook: I Tested the Defender Response for AI Agent Attacks
Review update (September 26, 2026): The source reviewed September 26 requires explicit GA model/version selection and Python 3.12 hash-locked dependencies. Current rules use stable alert IDs and ingestion freshness, without a fixed artificial …
Block Prompt Injection at the Network Layer with Microsoft Entra Prompt Injection Protection
Review update (September 25, 2026): Generative AI Insights remains preview and can retain full prompts with user identity. Set access, retention and user-notice policy before enabling it. Portal screenshots below are retained from the March 2026 …
Just-In-Time Access for AI Agents: Building a ZSP Gateway in Azure
Review update (September 25, 2026): The merged September 25 source rejects ambiguous/dot-segment scopes and mismatched ARM response identities, uses explicit managed identity with closed SDK clients, and moves synchronous operations off the async …
Building an LLM Prompt Injection Firewall with AWS Lambda
Review update (September 25, 2026): The September 25 merged source uses documented metric-filter IS TRUE/IS FALSE operators while retaining the distinct Logs Insights grammar. Metrics, alarm and dashboard share a project-specific namespace. New …
Sentinel MCP Server: Securing Your SOC's New AI Attack Surface
Review update (September 25, 2026): The walkthrough captures show January 6, 2026 evidence and an earlier tool inventory. The related detection lab is a source-reviewed identity/activity lab; it does not by itself prove ingestion or coverage of …
Securing the Agentic Workforce: Microsoft's Zero Trust for AI Agents
Status update, August 13, 2026: Microsoft Agent 365 is generally available for commercial customers, effective May 1, 2026, and the Microsoft Entra Agent ID platform reached GA in April 2026. The Ignite 2025 rollout history below remains useful …