AWS Security ChangesHomeSearch

AWS wellarchitected: Add AGENTSEC07 human-oversight/rogue-agent best-practice links

Service: wellarchitected · 2026-10-20 · Documentation medium

File: wellarchitected/latest/agentic-ai-lens/agentsec07.md

Summary

Adds five links to AGENTSEC07 best-practice pages on cognitive load management, confidence indicators and manipulation warnings, multiple reviewers, anomaly detection/containment, and red teaming.

Security assessment

The links document protections for human oversight and detection of rogue agents, including red teaming and containment — security guidance, but no specific vulnerability or advisory is cited in the diff.

Evidence

+[AGENTSEC07-BP05 Regular security assessments and red teaming](./agentsec07-bp05.html)  

Diff

diff --git a/wellarchitected/latest/agentic-ai-lens/agentsec07.md b/wellarchitected/latest/agentic-ai-lens/agentsec07.md
index 37a8ef9..e5fba52 100644
--- a//wellarchitected/latest/agentic-ai-lens/agentsec07.md
+++ b//wellarchitected/latest/agentic-ai-lens/agentsec07.md
@@ -14,0 +15,5 @@ AGENTSEC07: How do you protect human oversight from manipulation and detect rogu
+[AGENTSEC07-BP01 Implement cognitive load management](./agentsec07-bp01.html)  
+[AGENTSEC07-BP02 Clear confidence indicators and manipulation warnings](./agentsec07-bp02.html)  
+[AGENTSEC07-BP03 Multiple reviewers for critical operations](./agentsec07-bp03.html)  
+[AGENTSEC07-BP04 Behavioral anomaly detection and agent containment](./agentsec07-bp04.html)  
+[AGENTSEC07-BP05 Regular security assessments and red teaming](./agentsec07-bp05.html)