Microsoft Certified: Azure AI Fundamentals (AI-901)

Which two skill domains make up the Microsoft AI-901 exam blueprint?

• AIdentify AI concepts and responsibilities, and Implement AI solutions by using Microsoft…
• BMachine learning operations, and Data engineering on Azure
• CDesigning Azure infrastructure, and Securing Azure networks
• DPower Platform automation, and Microsoft 365 administration

What is the unified Microsoft platform for building and deploying AI applications and agents that AI-901 is built around?

• AMicrosoft Foundry
• BMicrosoft Power BI
• CMicrosoft Sentinel
• DMicrosoft Defender

Which responsible AI principle is most directly concerned with ensuring an AI system treats all groups of people equitably?

• AFairness
• BTransparency
• CAccountability
• DReliability and safety

A bank wants to make sure its loan-approval AI model does not disadvantage applicants based on ethnicity. Which responsible AI principle does this address?

• AFairness
• BInclusiveness
• CPrivacy and security
• DTransparency

Which responsible AI principle requires that AI systems should perform consistently and operate safely even in unexpected conditions?

• AReliability and safety
• BInclusiveness
• CAccountability
• DFairness

An AI system must protect personal data and give users control over how their information is used. Which responsible AI principle applies?

• APrivacy and security
• BTransparency
• CReliability and safety
• DAccountability

Designing an AI solution so that it is accessible to people with disabilities best reflects which responsible AI principle?

• AInclusiveness
• BFairness
• CTransparency
• DPrivacy and security

Users should understand how an AI system reaches its decisions and what its limitations are. Which principle does this describe?

• ATransparency
• BAccountability
• CInclusiveness
• DReliability and safety

Which responsible AI principle states that people, not the AI system itself, must be answerable for how an AI system operates?

• AAccountability
• BTransparency
• CFairness
• DPrivacy and security

A content team wants to block a generative AI app from producing hateful or violent text. Which Azure capability is designed for this?

• AAzure AI Content Safety
• BAzure Monitor
• CAzure Key Vault
• DAzure Resource Manager

Which statement best describes how a generative AI language model produces its output?

• AIt predicts the next token in a sequence based on patterns learned from large amounts of…
• BIt looks up the exact answer from a fixed database of facts
• CIt runs a fixed if-then rule set written by engineers
• DIt searches the live web for each response in real time by default

What is a token in the context of large language models?

• AA chunk of text, such as a word or part of a word, that the model processes
• BA security credential used to authenticate to the model
• CA billing unit equal to one full conversation
• DA pixel block used in image generation

Which type of model creates a numeric vector representation of text so that similar meanings are placed close together?

• AAn embedding model
• BA speech synthesis model
• CA reinforcement-only model
• DA rule-based parser

What does the temperature parameter control when calling a generative language model?

• AThe randomness and creativity of the generated output
• BThe maximum number of input documents allowed
• CThe physical CPU temperature of the host
• DThe number of users who can call the model

Which parameter limits how long a model's generated response can be?

• AMax tokens
• BTemperature
• CTop-p
• DFrequency penalty