Microsoft Azure AI Fundamentals Practice Exam 2025: Latest Questions

A retail company wants to reduce the risk of biased hiring decisions when using an AI model to screen resumes. Which action is most appropriate to address responsible AI considerations?

You need an Azure service that lets you build, train, and manage machine learning models using notebooks and automated ML, and also supports model deployment. Which service should you use?

A team wants to predict next month’s electricity usage (a numeric value) based on historical meter readings and weather data. What type of machine learning problem is this?

A manufacturing plant wants to detect whether a product is defective using images labeled as "defective" or "non-defective." What is the expected output of the model?

A customer support system must analyze incoming messages and route them to Billing, Technical, or Sales. Which NLP capability is the best fit?

You want to extract entities such as people, locations, and organizations from unstructured text like emails. Which Azure capability should you use?

A company wants to build a chatbot that answers questions using content from internal policy documents, and it must cite sources to reduce hallucinations. Which architecture pattern best fits?

A data scientist trained a model in Azure Machine Learning and wants to track metrics, datasets, and versions so results can be reproduced later. Which feature supports this requirement?

You are using a large language model to generate customer-facing email drafts. The business requires reducing harmful or inappropriate content while still allowing helpful responses. What is the best approach?

A computer vision model performs well in the lab but fails after deployment because lighting conditions in the factory vary significantly from the training images. What is the most likely issue and best corrective action?

You want to analyze customer comments and automatically assign a positive, negative, or neutral label. Which NLP workload is this?

You need to predict whether a loan applicant will default (Yes/No) based on historical application data. Which machine learning task is most appropriate?

A support team wants a chatbot to answer questions using an internal policy document. The bot must base answers on that document content. Which approach best meets this requirement?

Your team trained a model that performs well on the training dataset but performs poorly on new data from production. Which issue is most likely?

You are building an app that must identify the location of products on a warehouse shelf and draw bounding boxes around each item in an image. Which computer vision capability is required?

A call center wants to transcribe live customer calls into text and then summarize the conversation. Which combination of Azure AI capabilities best fits?

You plan to deploy an ML model to help with hiring decisions. Which action is a best practice to support responsible AI?

A team wants to build a model in Azure Machine Learning and keep track of model versions, training runs, and associated metrics so they can reproduce results later. Which feature is most relevant?

You deployed a generative AI assistant for employees. During testing, some users are able to prompt the model to reveal system instructions and internal configuration details. What should you implement to mitigate this risk?

A company trained a model using labeled images captured in a bright warehouse. After deployment to a dimly lit warehouse, accuracy drops significantly. What is the most likely cause?

A support team wants an AI system to automatically route incoming customer emails to the correct department (Billing, Technical, Sales). They have historical emails labeled with the correct department. Which type of machine learning is most appropriate?

You build an Azure AI Language model to extract key phrases from product reviews. After deployment, performance drops because customers start using new slang and product names not seen during development. What is the most appropriate action to improve results over time?

A company wants to quickly build a proof-of-concept chatbot that answers questions using content from their internal policy documents. They want to minimize custom model training and focus on grounding answers in their documents. Which approach best fits this requirement?

A retail store wants to analyze in-store camera footage to count how many people enter the store each hour and to track foot traffic patterns. Which computer vision workload does this represent?

You are designing an AI solution that reviews loan applications. Regulators require the company to explain why an application was approved or denied and to identify potential bias in the decision process. Which principle should be prioritized in the AI system design?