AI-900 Practice Question: Describe features of generative AI workloads on Azure
A developer uses Azure OpenAI to generate customer support responses. The developer wants to ensure that the model does not produce responses that contain offensive, hateful, or harmful language, even when users input problematic prompts. Which Azure OpenAI feature should the developer configure to achieve this?
⚠ Common exam trap
Watch out — candidates often confuse content filtering with model tuning parameters like temperature or frequency penalty, assuming that adjusting output randomness or repetition can prevent harmful content, when in fact only a dedicated content filter can enforce safety policies.
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
✓
Enabling the content filter
The content filter in Azure OpenAI is specifically designed to detect and block offensive, hateful, or harmful language in both user prompts and model responses. By enabling this feature, the developer ensures that even if a user submits a problematic input, the model's output will be filtered to prevent generating inappropriate content. This directly addresses the requirement to avoid harmful language.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Setting a low temperature value
Why it's wrong here
Temperature governs the randomness of token sampling by scaling the probability distribution before sampling, so a low value makes completions more predictable and greedy. It cannot evaluate whether a phrase is offensive; the model's pretrained priors still include toxic patterns, and those can surface even with deterministic decoding. Content safety requires an explicit moderation layer, not a sampling parameter.
- ✗
Limiting the max_tokens parameter
Why it's wrong here
max_tokens is a hard limit on the number of tokens returned, and raising or lowering it only changes where generation stops. Harmful content can appear in the very first few tokens, so a smaller token budget does not prevent an offensive answer; it merely truncates it before or after the banned phrase. This parameter controls output length only, not semantic filtering.
- ✓
Enabling the content filter
Why this is correct
Enabling the Azure OpenAI content filter activates the built-in moderation pipeline, which uses trained classifiers to score prompts and completions across hate, violence, sexual content, self-harm, and other categories, and blocks or masks content above a configured severity threshold. This is the designed mechanism for meeting safety requirements because it inspects actual output content, not just generation behavior. The filter can be configured with differing severity levels for both input and output.
- ✗
Setting a high frequency penalty
Why it's wrong here
Frequency penalty is a repetition control that subtracts a penalty from token logits each time a token has already appeared in the generated sequence, thereby encouraging lexical diversity. It has no understanding of whether a token sequence is harmful, and in fact increasing diversity could introduce different toxic phrases that would not otherwise have repeated. Such a penalty can neither block nor flag disallowed content.
Go deeper
Related to this question
Learn chapter
Azure Machine Learning Studio
Key term
Feature
A feature is a distinct unit of functionality that delivers value to the user, often managed and tracked throughout the software development lifecycle.
Key term
Content filter
A content filter is an AI-powered safety system that screens user prompts and AI-generated responses for harmful, offensive, or restricted content, helping to ensure responsible use of Azure AI services.
About these practice questions
Courseiva writes every AI-900 question from scratch — 985 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. 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.