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 5

Question 40

An Azure AI solution uses the following approved control: Run the required text analysis operations separately and preserve distinct result types. What observable result should pass QA?

  1. The generated audio plays through the selected output device
  2. The response contains both an attitude label and typed entity spans
  3. The output audio reads the approved sentence intelligibly
  4. Mic capture ceases and the app no longer receives audio frames

Explanation

Correct Answer

B:

Sentiment polarity and named entity spans answer different questions. The expected acceptance evidence is: The response contains both an attitude label and typed entity spans. This observation directly tests the stated behavior.

Incorrect Answers

A:

Successful synthesis does not guarantee local audio is routed to a speaker. This observation does not establish the required result: The response contains both an attitude label and typed entity spans.

C:

Speech synthesis generates audio waveforms from supplied text. This observation does not establish the required result: The response contains both an attitude label and typed entity spans.

D:

User control of capture must be enforced at the device or application layer. This observation does not establish the required result: The response contains both an attitude label and typed entity spans.

Question 41

The approved design action is: Limit retrieval and chat context to the authenticated case scope. Which test result would BEST establish that it worked?

  1. An authenticated test request returns a response from the chosen deployment
  2. A representative maximum-size test stays within documented budgets
  3. An account-specific summary excludes cross-account conversation data
  4. Conflict tests keep disallowed actions blocked

Explanation

Correct Answer

C:

Source scoping enforces privacy and also improves answer relevance. The expected acceptance evidence is: An account-specific summary excludes cross-account conversation data. This observation directly tests the stated behavior.

Incorrect Answers

A:

The portal provides model selection, deployment and test interfaces for supported resources. This observation does not establish the required result: An account-specific summary excludes cross-account conversation data.

B:

Capabilities and limits vary by model deployment. This observation does not establish the required result: An account-specific summary excludes cross-account conversation data.

D:

Developer and system constraints govern allowed operations over user requests. This observation does not establish the required result: An account-specific summary excludes cross-account conversation data.

Question 42

An acceptance tester needs to validate: Use digit-focused recognition hints and explicit confirmation for consequential IDs. Which observed result proves the intended behavior?

  1. A spoken inquiry yields the correct audible answer after processing
  2. The caller confirms the extracted account number before lookup
  3. The model's response reflects the words spoken in the actual recording
  4. The listener hears equivalent French instructions

Explanation

Correct Answer

B:

Speech recognition uncertainty is higher for confusable numeric strings. The expected acceptance evidence is: The caller confirms the extracted account number before lookup. This observation directly tests the stated behavior.

Incorrect Answers

A:

Input transcription and output synthesis are distinct steps. This observation does not establish the required result: The caller confirms the extracted account number before lookup.

C:

A text-only request cannot interpret sound content by file name alone. This observation does not establish the required result: The caller confirms the extracted account number before lookup.

D:

Cross-language audio output requires both language transfer and synthesis. This observation does not establish the required result: The caller confirms the extracted account number before lookup.

Question 43

An AI operations team is testing two distinct capabilities. Failure (1): A generated answer invents a page number with no source anchor. Failure (2): The checkbox is discarded during plain text extraction. Which TWO specific corrections should the team apply?

  1. Keep transcript anchors and validate extracted tasks
  2. Include selection-mark processing in the analyzer
  3. Redact logs and use access-controlled short-lived source storage
  4. Enable appropriate diarization and handle ambiguous overlap
  5. Preserve actual section/page references through retrieval

Explanation

Correct Answers

B:

Correct for failure (2). Checkbox state is a structured signal rather than arbitrary prose. It independently corrects the second defect.

E:

Correct for failure (1). Structured layout and provenance let users verify a quoted clause. The proposed correction directly addresses that reported defect.

Incorrect Answers

A:

Transcription and summarization are separate results with different validation. In this case the first outcome requires the team to use document extraction with source positions before indexing for search, and the second requires it to extract selection marks as explicit typed values. This option does not replace those two separate controls.

C:

Inference outputs and telemetry can contain protected identity information. In this case the first outcome requires the team to use document extraction with source positions before indexing for search, and the second requires it to extract selection marks as explicit typed values. This option does not replace those two separate controls.

D:

Conversation extraction needs more than a single flattened text stream. In this case the first outcome requires the team to use document extraction with source positions before indexing for search, and the second requires it to extract selection marks as explicit typed values. This option does not replace those two separate controls.

Question 44

An engineering review documents this requirement: A voice interface must recognize Spanish requests from customers. Select the most appropriate solution.

  1. Configure a supported Spanish recognition locale and test representative speakers
  2. Use a translation service or supported multilingual generation with evaluation
  3. Combine recognition or audio understanding with verified text and text-to-speech
  4. Use supported SSML phoneme or say-as controls for pronunciation

Explanation

Correct Answer

A:

Recognition models need the appropriate supported language settings. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Translation requires semantic equivalence and language direction. That is not sufficient here: Recognition models need the appropriate supported language settings.

C:

Input transcription and output synthesis are distinct steps. That is not sufficient here: Recognition models need the appropriate supported language settings.

D:

SSML can guide pronunciation beyond plain-text guessing. That is not sufficient here: Recognition models need the appropriate supported language settings.

Question 45

An acceptance tester needs to validate: Wait for the analysis operation to reach a terminal state before reading results. Which observed result proves the intended behavior?

  1. The result is read only after the operation succeeds
  2. Several labeled forms populate the same typed output fields
  3. A reply links its claim to a transcript time or relevant video segment
  4. The resulting record maps each value to its correct page

Explanation

Correct Answer

A:

Long-running media analysis may be asynchronous. The expected acceptance evidence is: The result is read only after the operation succeeds. This observation directly tests the stated behavior.

Incorrect Answers

B:

A schema-driven analyzer identifies business fields independent of fixed coordinates. This observation does not establish the required result: The result is read only after the operation succeeds.

C:

Multimodal grounding should preserve how the answer maps to source media. This observation does not establish the required result: The result is read only after the operation succeeds.

D:

Page order and references are part of document provenance. This observation does not establish the required result: The result is read only after the operation succeeds.

Question 46

An acceptance tester needs to validate: Use an agent orchestration plan with separate approval before booking. Which observed result proves the intended behavior?

  1. Names in a labeled sample are detected and redacted
  2. The product shape remains unchanged against the new background
  3. The narrated explanation matches labeled chart features without inventing values
  4. The itinerary can be prepared without a charge until confirmation

Explanation

Correct Answer

D:

An agent may propose actions but payment must remain permission-controlled. The expected acceptance evidence is: The itinerary can be prepared without a charge until confirmation. This observation directly tests the stated behavior.

Incorrect Answers

A:

Entity extraction locates sensitive spans for privacy handling. This observation does not establish the required result: The itinerary can be prepared without a charge until confirmation.

B:

Editing changes selected pixels while leaving product regions intact. This observation does not establish the required result: The itinerary can be prepared without a charge until confirmation.

C:

Multimodal input can describe the graphic and speech output can read that description. This observation does not establish the required result: The itinerary can be prepared without a charge until confirmation.

Question 47

Troubleshooting report: A model endpoint is regional but telemetry goes to an unapproved location. Which correction is MOST directly targeted to this failure?

  1. Audit and correct all cross-region ancillary data flows
  2. Run native-language acceptance cases with verified translations and experts
  3. Add post-retrieval and pre-action policy enforcement
  4. Rebalance safety tests and adjust erroneous dialect associations

Explanation

Correct Answer

A:

This correction directly addresses the reported failure. Privacy and compliance require controlling the full data path. The targeted change is: Audit and correct all cross-region ancillary data flows.

Incorrect Answers

B:

Inclusion requires equivalent safety and meaning across supported locales. In the stated case, troubleshooting report: A model endpoint is regional but telemetry goes to an unapproved location; “Run native-language acceptance cases with verified translations and experts” does not target that failure.

C:

Threats can enter later stages even when the original user request was harmless. In the stated case, troubleshooting report: A model endpoint is regional but telemetry goes to an unapproved location; “Add post-retrieval and pre-action policy enforcement” does not target that failure.

D:

Fairness requires discriminating actual harm from linguistic variety. In the stated case, troubleshooting report: A model endpoint is regional but telemetry goes to an unapproved location; “Rebalance safety tests and adjust erroneous dialect associations” does not target that failure.

Question 48

An AI operations team is testing two distinct capabilities. Failure (1): Large requests fail due to unbounded prompt content. Failure (2): Only the final prose is logged with no tool execution details. Which TWO specific corrections should the team apply?

  1. Constrain request size and trim or chunk input before sending
  2. Enable privacy-minimized structured agent tracing
  3. Freeze representative evaluation cases and compare versions
  4. Move the model call behind a secured server endpoint
  5. Validate case identity and entitlement before each tool call

Explanation

Correct Answers

A:

Correct for failure (1). Capabilities and limits vary by model deployment. The proposed correction directly addresses that reported defect.

B:

Correct for failure (2). Operational observability aids debugging without exposing hidden sensitive prompt data. It independently corrects the second defect.

Incorrect Answers

C:

Anecdotal success on different prompts cannot establish an improvement. In this case the first outcome requires the team to check supported context and output limits for the chosen deployment, and the second requires it to capture allowed tool calls, response statuses and source references in traces. This option does not replace those two separate controls.

D:

Backend controls protect credentials, quotas and data policy. In this case the first outcome requires the team to check supported context and output limits for the chosen deployment, and the second requires it to capture allowed tool calls, response statuses and source references in traces. This option does not replace those two separate controls.

E:

Structured state prevents subtle substitution of unrelated customer records. In this case the first outcome requires the team to check supported context and output limits for the chosen deployment, and the second requires it to capture allowed tool calls, response statuses and source references in traces. This option does not replace those two separate controls.

Question 49

An Azure AI solution uses the following approved control: Extract selection marks as explicit typed values. What observable result should pass QA?

  1. The output identifies both amount and confirmed currency
  2. The output returns correct limits with supporting clauses
  3. The checked option is returned accurately as true or selected
  4. Observed labeled events receive evidence-backed intervals and brief unobserved events are flagged for review

Explanation

Correct Answer

C:

Checkbox state is a structured signal rather than arbitrary prose. The expected acceptance evidence is: The checked option is returned accurately as true or selected. This observation directly tests the stated behavior.

Incorrect Answers

A:

A currency symbol or amount alone may not establish financial meaning. This observation does not establish the required result: The checked option is returned accurately as true or selected.

B:

Custom extraction maps semantically relevant phrases to typed fields. This observation does not establish the required result: The checked option is returned accurately as true or selected.

D:

Temporal evidence supports the timing of observed events, but approximately one frame per second cannot guarantee subsecond event capture. This observation does not establish the required result: The checked option is returned accurately as true or selected.

Question 50

A project needs the following AI behavior: An architect wants an AI model to classify a picture as indoors or outdoors. What should the designer implement?

  1. Use regression or forecasting appropriate to time-series data
  2. Use object detection when each product needs a location
  3. Use image classification for the requested category
  4. Use a generative language workload with policy-grounded response constraints

Explanation

Correct Answer

C:

Image classification predicts image-level labels. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Forecasts predict future numeric quantities rather than discrete tags. That is not sufficient here: Image classification predicts image-level labels.

B:

Object detectors return classes and spatial boundaries. That is not sufficient here: Image classification predicts image-level labels.

D:

Generation produces novel text, unlike choosing a fixed class label. That is not sufficient here: Image classification predicts image-level labels.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests