AI-900 Practice Question: Describe features of generative AI workloads on Azure
A company uses Azure OpenAI Service to generate marketing copy. They want to ensure that the generated text does not contain offensive language or harmful stereotypes, even if the prompt inadvertently leads the model in that direction. Which Azure OpenAI feature should they configure to help prevent such outputs?
⚠ Common exam trap
Many exam-takers confuse content filtering with prompt engineering, assuming that careful prompt design alone can prevent harmful outputs, but Azure OpenAI's content filtering is the dedicated safety mechanism that operates independently of prompt quality.
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 filtering
Content filtering in Azure OpenAI Service uses a set of pre-built, configurable filters to detect and block harmful content categories such as hate, violence, sexual, and self-harm. This feature operates at the service level, intercepting both prompts and completions to prevent offensive language or harmful stereotypes from being generated, regardless of how the prompt is phrased.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Content filtering
Why this is correct
Content filtering in Azure OpenAI Service is an integrated safety layer that evaluates both input prompts and output completions against configurable severity thresholds for categories such as hate, sexual, violence, and self-harm. It prevents offensive or harmful marketing copy from being delivered even if the model unintentionally produces it, because filtering operates independently of prompt phrasing and is enforced as part of the service's generation pipeline. This makes it the correct control for ensuring output safety.
- ✗
Prompt engineering
Why it's wrong here
Prompt engineering involves carefully crafting the instruction, context, or constraints in the input that steer the model's style and focus, such as asking for 'professional, friendly ad copy,' but it is not a governance or filtering mechanism. While an engineered prompt can instruct the model to avoid certain words, the model may still produce harmful text if the underlying generation path drifts, and there is no enforcement process to block it. The only guarantee comes from a dedicated safety filter, so prompt engineering alone is not sufficient.
- ✗
Fine-tuning
Why it's wrong here
Fine-tuning retrains the base model on a curated dataset, adjusting weights to match a specific brand voice, audience, or marketing domain, but it does not introduce a moderation layer or output validation step. Even after fine-tuning, the model's original capabilities and latent biases remain, and the fine-tuned model can generate offensive language if the training data contains patterns or if prompted in an adversarial way. Content filtering must still be applied externally or via the Azure OpenAI service to catch such outputs, making fine-tuning irrelevant as a safety guarantee.
- ✗
Few-shot learning
Why it's wrong here
Few-shot learning conditions the model by including several labeled examples of the desired output format in the prompt, helping it mimic tone or structure from sample marketing copy. This setting establishes context and patterns, but it functions only as input-based guidance; it cannot detect or suppress a toxic output that deviates from the samples. Because the examples are not rules and are not enforced by any post-generation mechanism, few-shot learning leaves harmful content unblocked and is therefore not a control for offensive language.
Go deeper
Related to this question
Learn chapter
Azure Machine Learning Studio
Key term
Azure OpenAI Service
Azure OpenAI Service is a cloud platform from Microsoft that lets developers use powerful artificial intelligence models, like GPT-4, to build applications that can understand and generate human-like text, code, images, and more.
Key term
Model
In IT and AI, a model is a trained mathematical representation that learns patterns from data to make predictions or decisions.
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.