AI-900 Practice Question: Describe features of generative AI workloads on Azure
A company uses Azure OpenAI to generate marketing copy. They want to ensure that the generated text does not contain inappropriate or harmful content before it is published. Which Azure OpenAI feature is specifically designed for this purpose?
⚠ Common exam trap
It's easy for candidates to confuse prompt engineering features (temperature, top-p, system message) with built-in safety mechanisms, assuming they can prevent harmful content when only content filters provide a deterministic, policy-enforced block.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Content filters
Content filters are the Azure OpenAI feature specifically designed to detect and block inappropriate or harmful content in generated text. They apply configurable severity levels across categories like hate, violence, self-harm, and sexual content, ensuring outputs meet safety policies before publication.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Temperature
Why it's wrong here
Temperature is a sampling-control parameter, not a safety mechanism. It modifies the probability distribution over next tokens by scaling logits before the softmax is applied; a higher value makes outputs more random and diverse, while a lower value makes them more deterministic and repetitive. Because it only changes which tokens are likely to be selected, it does not detect, block, or filter harmful content—unsafe text can still be sampled even at low temperature if it is plausible in the model's distribution.
- ✗
Top-p (nucleus sampling)
Why it's wrong here
Top-p (nucleus sampling) controls generation variety by restricting token selection to the smallest set whose cumulative probability reaches a threshold p. This influences the coherence and creativity of marketing copy, but it does not evaluate semantic meaning or content policy. An offensive or unsafe token can easily fall inside the chosen nucleus, so top-p provides no safeguarding against inappropriate outputs.
- ✗
System message
Why it's wrong here
A system message shapes the assistant's role, tone, and behavioral guidelines, such as "write catchy yet professional marketing copy." However, these instructions are advisory: the model may comply imperfectly, and they can be overridden by prompt injection or jailbreak attempts. In Azure OpenAI, system messages are part of the model context and do not perform deterministic filtering of generated text; content policy enforcement remains the job of dedicated content filters.
- ✓
Content filters
Why this is correct
Content filters are the correct answer because Azure OpenAI applies dedicated safety models that automatically screen prompts and completions for harmful content in categories such as hate, sexual content, violence, and self-harm. These filters assess severity levels and can block or annotate inappropriate content before it is returned. Unlike generation parameters or instructions, they are an actual enforcement layer designed to prevent harmful outputs in Azure OpenAI service.
Go deeper
Related to this question
About these practice questions
One of 985 original AI-900 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.