Agentic AI Reference Architecture

Safety & Security · Software component

Sensitive Topic Detector

Software componentSafety & SecuritySafety, Security & Governancearc:SensitiveTopicDetector

A classifier that detects conversation content on sensitive topics, such as harassment complaints, legal disputes, emotional distress or security concerns, that requires immediate human handling.

Responsibility. Flags sensitive-topic turns for immediate escalation regardless of system confidence.

is invoked bysends data toDialogue Flow Manager: is invoked byDialogue Flow ManagerEscalation Agent: sends data toEscalation Agent
Direct neighbourhood (hover for relationship types)

Relationships

is invoked by dependency

sends data to dynamic

Design guidance

Classification

Patterns
Topic-triggered escalation
Quality attributes
Safety (ISO/IEC 25010 | NIST AI RMF: safe)Interaction capability (ISO/IEC 25010)
Risks mitigated
AI handling situations that demand human empathy and judgment

Sources

  1. Ch10.1: T. Nguyen, "Conversational UI," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 10.1. ISBN: 9798244538229.