AI-900Domain 5: Describe Features of Generative AI Workloads on AzureGenerative AImediumSingle choice
Before showing AI-generated responses to users, an app must filter harmful categories — hate, violence, self-harm, sexual content — in both prompts and completions. Which Azure service provides this?
A.Microsoft Defender for Cloud
Defender is cloud security posture — not content moderation.
B.Azure AI Content Safety
Content Safety = the harm-filtering service. Correct.
C.Azure AI Language sentiment analysis
Sentiment ≠ safety — a toxic prompt can have 'neutral' sentiment.
D.Azure Policy
Azure Policy governs resource configurations — unrelated.
Why this answer
Azure AI Content Safety detects harmful content across text and images with severity scoring and customizable filters — the responsible-AI guardrail service for GenAI apps. Sentiment analysis measures opinion polarity, not harm categories.
Correct answer: B · Last reviewed 2026-02-14
Official AWS references
Practice this interactively
Answer without seeing the solution first in the Generative AI practice session.
Practice Generative AI