McAfee Secure

Exam Code: AI-901

Exam Name: Microsoft Azure AI Fundamentals

Certification Provider: Microsoft

Corresponding Certification: Microsoft Certified: Azure AI Fundamentals

Microsoft AI-901 Questions & Answers

Study with Up-To-Date REAL Exam Questions and Answers from the ACTUAL Test

166 Questions & Answers with Testing Engine

"Microsoft Azure AI Fundamentals Exam", also known as AI-901 exam, is a Microsoft certification exam.

Pass your tests with the always up-to-date AI-901 Exam Engine. Your AI-901 training materials keep you at the head of the pack!

Satisfaction Guaranteed

Money Back Guarantee

Testking provides hassle-free money back guarantee with our products. That is because we have 100% trust in the abilities of our professional and experience product team, and our record is a proof of that.

99.6% PASS RATE
Was: $137.49
Now: $124.99

Microsoft AI-901 Practice Test 17

Question 160

A project needs the following AI behavior: A Python client must call a deployed Foundry model rather than a model catalog entry. What should the designer implement?

  1. Configure the client with the actual resource endpoint and deployment identifier
  2. Check supported context and output limits for the chosen deployment
  3. Update the configured deployment name and verify compatibility with the SDK
  4. Treat user content as untrusted and maintain privileged instruction boundaries

Explanation

Correct Answer

A:

An inferencing request is routed to a concrete deployment configured in the resource. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Capabilities and limits vary by model deployment. That is not sufficient here: An inferencing request is routed to a concrete deployment configured in the resource.

C:

A model family label is not necessarily the concrete deployed endpoint. That is not sufficient here: An inferencing request is routed to a concrete deployment configured in the resource.

D:

A user cannot use text to elevate permissions beyond the application policy. That is not sufficient here: An inferencing request is routed to a concrete deployment configured in the resource.

Question 161

An engineering review documents this requirement: A journalist needs product and organization names from news articles. Select the most appropriate solution.

  1. Use layout-aware extraction and selection-mark analysis
  2. Use sentiment analysis to estimate positive, negative, or neutral tone
  3. Use text summarization with fidelity checks against sources
  4. Use named entity recognition with entity types

Explanation

Correct Answer

D:

NER locates and categorizes named entities in text. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Different form elements require more than ordinary running-text OCR. That is not sufficient here: NER locates and categorizes named entities in text.

B:

Sentiment analysis infers expressed attitude rather than extracting names. That is not sufficient here: NER locates and categorizes named entities in text.

C:

Summarization compresses content without independently validating missing facts. That is not sufficient here: NER locates and categorizes named entities in text.

Question 162

An acceptance tester needs to validate: Verify the service credential, resource and authorized region. Which observed result proves the intended behavior?

  1. The translated request preserves numbers and intent
  2. The recognizer succeeds with valid resource credentials
  3. The transcript matches a labeled spoken test phrase
  4. Labeled specialty phrases are transcribed with reduced errors

Explanation

Correct Answer

B:

Audio capture success does not establish remote service authentication. The expected acceptance evidence is: The recognizer succeeds with valid resource credentials. This observation directly tests the stated behavior.

Incorrect Answers

A:

Translation requires semantic equivalence and language direction. This observation does not establish the required result: The recognizer succeeds with valid resource credentials.

C:

Speech recognition converts waveform features to text tokens. This observation does not establish the required result: The recognizer succeeds with valid resource credentials.

D:

Domain-specific words may require biasing or adaptation beyond a generic recognizer. This observation does not establish the required result: The recognizer succeeds with valid resource credentials.

Question 163

A solution team has selected this control: Set explicit uncertainty thresholds and route high-risk decisions to qualified review. Which observation is the most direct acceptance evidence?

  1. A simulated timeout yields a failure status without a fabricated numeric balance
  2. Missing dosage context triggers abstention and a review route
  3. A compliance record links each exception to an owner and disposition
  4. An unprivileged test user is denied protected operations despite claiming to be an admin

Explanation

Correct Answer

B:

Safety-critical recommendations require verified constraints outside open-ended generation. The expected acceptance evidence is: Missing dosage context triggers abstention and a review route. This observation directly tests the stated behavior.

Incorrect Answers

A:

A fluent generation cannot substitute for missing authoritative transaction data. This observation does not establish the required result: Missing dosage context triggers abstention and a review route.

C:

Self-attestation is not independent proof of regulatory controls. This observation does not establish the required result: Missing dosage context triggers abstention and a review route.

D:

Natural-language claims cannot elevate authentication authority. This observation does not establish the required result: Missing dosage context triggers abstention and a review route.

Question 164

An Azure AI Fundamentals solution has this requirement: A voice assistant must switch between approved male and female synthetic voice styles. Which implementation most directly addresses it?

  1. Evaluate custom phrase hints or supported speech customization on representative terms
  2. Apply approved audio retention and access rules and keep only required derived text
  3. Select a supported synthesis voice ID rather than altering transcript text
  4. Use speech-to-text recognition on a supported audio input stream

Explanation

Correct Answer

C:

Voice timbre is a synthesis configuration, not speech recognition output. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Domain-specific words may require biasing or adaptation beyond a generic recognizer. That is not sufficient here: Voice timbre is a synthesis configuration, not speech recognition output.

B:

Speech workloads still process personal data and require controlled lifecycle. That is not sufficient here: Voice timbre is a synthesis configuration, not speech recognition output.

D:

Speech recognition converts waveform features to text tokens. That is not sufficient here: Voice timbre is a synthesis configuration, not speech recognition output.

Question 165

An engineering review documents this requirement: A kiosk must read text instructions aloud for visually impaired visitors. Select the most appropriate solution.

  1. Use text-to-speech with a supported voice and accessible controls
  2. Use visual-question answering with a multimodal model
  3. Use speech-to-text transcription before indexing call content
  4. Use semantic document indexing and retrieval rather than only keyword equality

Explanation

Correct Answer

A:

Speech synthesis converts supplied text into audible output. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Image-and-text reasoning can connect a question with visible scene content. That is not sufficient here: Speech synthesis converts supplied text into audible output.

C:

Speech recognition converts speech into words suitable for search. That is not sufficient here: Speech synthesis converts supplied text into audible output.

D:

Semantic search can retrieve conceptually related passages. That is not sufficient here: Speech synthesis converts supplied text into audible output.

Question 166

An AI operations team is testing two distinct capabilities. Failure (1): The system executes whenever an unstable partial result contains a keyword. Failure (2): The app selects a voice ID that is not supported in that locale. Which TWO specific corrections should the team apply?

  1. Gate side effects on final recognition and explicit confirmation
  2. Use a compatible voice from the service's voice catalog
  3. Add a no-speech branch and clarification option
  4. Keep operation outputs separate and map each to its intended field
  5. Correct streaming buffer and session lifecycle

Explanation

Correct Answers

A:

Correct for failure (1). Interim hypotheses can change as more audio arrives. The proposed correction directly addresses that reported defect.

B:

Correct for failure (2). Not every voice is supported for all languages. It independently corrects the second defect.

Incorrect Answers

C:

An empty recognition result is not semantically equivalent to the word no. In this case the first outcome requires the team to wait for supported final recognition results before high-impact tool calls, and the second requires it to check synthesized voice availability for the selected locale and region. This option does not replace those two separate controls.

D:

Sentiment polarity and named entity spans answer different questions. In this case the first outcome requires the team to wait for supported final recognition results before high-impact tool calls, and the second requires it to check synthesized voice availability for the selected locale and region. This option does not replace those two separate controls.

E:

Real-time audio pipelines need coordinated capture and recognition start. In this case the first outcome requires the team to wait for supported final recognition results before high-impact tool calls, and the second requires it to check synthesized voice availability for the selected locale and region. This option does not replace those two separate controls.

Question 167

An Azure AI solution uses the following approved control: Select a supported synthesis voice ID rather than altering transcript text. What observable result should pass QA?

  1. The acronym is spoken according to the approved pronunciation sample
  2. Two voice selections read identical text with different timbre
  3. A test email returns correct product and organization spans
  4. The translated request preserves numbers and intent

Explanation

Correct Answer

B:

Voice timbre is a synthesis configuration, not speech recognition output. The expected acceptance evidence is: Two voice selections read identical text with different timbre. This observation directly tests the stated behavior.

Incorrect Answers

A:

SSML can guide pronunciation beyond plain-text guessing. This observation does not establish the required result: Two voice selections read identical text with different timbre.

C:

NER identifies named spans rather than summarizing the email. This observation does not establish the required result: Two voice selections read identical text with different timbre.

D:

Translation requires semantic equivalence and language direction. This observation does not establish the required result: Two voice selections read identical text with different timbre.

Question 168

Troubleshooting report: A tool executes merely because a user typed please proceed. Which correction is MOST directly targeted to this failure?

  1. Register the tool and validate its description and argument schema
  2. Reinforce role boundaries and enforce secrets outside prompts
  3. Attach idempotency keys and verify tool receipts after reconnecting
  4. Add approval state and verified role checks

Explanation

Correct Answer

D:

This correction directly addresses the reported failure. High-impact side effects need permissions independent of conversational language. The targeted change is: Add approval state and verified role checks.

Incorrect Answers

A:

Tools must be discoverable with valid schemas before selection. In the stated case, troubleshooting report: A tool executes merely because a user typed please proceed; “Register the tool and validate its description and argument schema” does not target that failure.

B:

Developer and system constraints govern allowed operations over user requests. In the stated case, troubleshooting report: A tool executes merely because a user typed please proceed; “Reinforce role boundaries and enforce secrets outside prompts” does not target that failure.

C:

Client retries should distinguish read requests from side-effecting tool calls. In the stated case, troubleshooting report: A tool executes merely because a user typed please proceed; “Attach idempotency keys and verify tool receipts after reconnecting” does not target that failure.

Question 169

An acceptance tester needs to validate: Use document information extraction with a typed field schema. Which observed result proves the intended behavior?

  1. Answers cite valid manual passages and avoid unsupported details
  2. Parsed fields match labeled form values and missing fields remain explicit
  3. Test feedback returns relevant phrases found in the message
  4. A prompt-only request yields a new original illustration rather than a classification label

Explanation

Correct Answer

B:

Document extraction maps unstructured layouts into structured values. The expected acceptance evidence is: Parsed fields match labeled form values and missing fields remain explicit. This observation directly tests the stated behavior.

Incorrect Answers

A:

RAG retrieves current external evidence for a generated response. This observation does not establish the required result: Parsed fields match labeled form values and missing fields remain explicit.

C:

Key phrase extraction finds prominent topics without inventing new source material. This observation does not establish the required result: Parsed fields match labeled form values and missing fields remain explicit.

D:

Text-to-image generation creates new visual content from a description; classification only assigns labels to existing inputs. This observation does not establish the required result: Parsed fields match labeled form values and missing fields remain explicit.

Question 170

An Azure AI Fundamentals solution has this requirement: A bot must answer questions about what is happening in an uploaded screenshot. Which implementation most directly addresses it?

  1. Use an agent workflow with a permitted ticket tool and an authorization gate
  2. Use object detection and count accepted object instances
  3. Use visual-question answering with a multimodal model
  4. Use named entity recognition with entity types

Explanation

Correct Answer

C:

Image-and-text reasoning can connect a question with visible scene content. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Agentic workloads combine reasoning and bounded external actions. That is not sufficient here: Image-and-text reasoning can connect a question with visible scene content.

B:

Instance detections support counts where one image contains several objects. That is not sufficient here: Image-and-text reasoning can connect a question with visible scene content.

D:

NER locates and categorizes named entities in text. That is not sufficient here: Image-and-text reasoning can connect a question with visible scene content.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests