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 42

Question 410

Troubleshooting report: The app provides one global category for several separate objects. Which correction is MOST directly targeted to this failure?

  1. Constrain the prompt and validate the final artwork
  2. Use detection instead of whole-image classification
  3. Switch to object detection for localized results
  4. Review multiple timestamped frames and flag rapid or unobserved steps instead of inferring them

Explanation

Correct Answer

C:

This correction directly addresses the reported failure. Classification predicts image-level categories while detection provides positions. The targeted change is: Switch to object detection for localized results.

Incorrect Answers

A:

Explicit visual guidance can influence generated appearance but is not absolute. In the stated case, troubleshooting report: The app provides one global category for several separate objects; “Constrain the prompt and validate the final artwork” does not target that failure.

B:

Object detection reports which objects appear and where they occur. In the stated case, troubleshooting report: The app provides one global category for several separate objects; “Use detection instead of whole-image classification” does not target that failure.

D:

A single still cannot convey action order; sampled video frames may also omit fleeting intermediate steps. In the stated case, troubleshooting report: The app provides one global category for several separate objects; “Review multiple timestamped frames and flag rapid or unobserved steps instead of inferring them” does not target that failure.

Question 411

A solution team has selected this control: Audit training labels and evaluate corrected outcome criteria. Which observation is the most direct acceptance evidence?

  1. The model/deployment capability check and a spoken-audio test confirm that this endpoint supports transcription
  2. The output is produced as a sequence of tokens conditioned on the prompt
  3. A permitted larger output budget allows the complete response
  4. Adjusted label evaluation measures outcomes under appropriate criteria

Explanation

Correct Answer

D:

Bias can enter through the targets themselves, not just inputs. The expected acceptance evidence is: Adjusted label evaluation measures outcomes under appropriate criteria. This observation directly tests the stated behavior.

Incorrect Answers

A:

Model capabilities must include the required audio modality before the client can obtain speech-to-text results. This observation does not establish the required result: Adjusted label evaluation measures outcomes under appropriate criteria.

B:

A generative language model estimates likely continuations rather than reading a stored answer paragraph. This observation does not establish the required result: Adjusted label evaluation measures outcomes under appropriate criteria.

C:

Generation stops when an output budget or stop condition is reached. This observation does not establish the required result: Adjusted label evaluation measures outcomes under appropriate criteria.

Question 412

Troubleshooting report: Audio recordings persist indefinitely in an unrestricted bucket. Which correction is MOST directly targeted to this failure?

  1. Apply data minimization, retention and permission controls
  2. Add auditable explanation records and an appeals workflow
  3. Run the frozen adversarial and ordinary acceptance suite against the candidate
  4. Permit calibrated caveats and qualified review paths

Explanation

Correct Answer

A:

This correction directly addresses the reported failure. Privacy requires meaningful notice and bounded handling of personal data. The targeted change is: Apply data minimization, retention and permission controls.

Incorrect Answers

B:

Accountability includes traceable responsibility and appropriate contestability. In the stated case, troubleshooting report: Audio recordings persist indefinitely in an unrestricted bucket; “Add auditable explanation records and an appeals workflow” does not target that failure.

C:

Reliability includes preventing behavior regression across model versions. In the stated case, troubleshooting report: Audio recordings persist indefinitely in an unrestricted bucket; “Run the frozen adversarial and ordinary acceptance suite against the candidate” does not target that failure.

D:

Transparency reduces mistaken reliance on unsupported professional guarantees. In the stated case, troubleshooting report: Audio recordings persist indefinitely in an unrestricted bucket; “Permit calibrated caveats and qualified review paths” does not target that failure.

Question 413

A solution team has selected this control: Use a supported diarization or speaker-labeling capability where available. Which observation is the most direct acceptance evidence?

  1. A confirmed final transcript triggers the intended action once
  2. Each action item maps to an utterance and assigned speaker
  3. A spoken inquiry yields the correct audible answer after processing
  4. The transcript marks each turn with a stable speaker label

Explanation

Correct Answer

D:

Speaker diarization separates turns and identities at a conversation level. The expected acceptance evidence is: The transcript marks each turn with a stable speaker label. This observation directly tests the stated behavior.

Incorrect Answers

A:

Interim hypotheses can change as more audio arrives. This observation does not establish the required result: The transcript marks each turn with a stable speaker label.

B:

Transcription obtains words; summarization extracts decisions and tasks. This observation does not establish the required result: The transcript marks each turn with a stable speaker label.

C:

Input transcription and output synthesis are distinct steps. This observation does not establish the required result: The transcript marks each turn with a stable speaker label.

Question 414

An AI operations team is testing two distinct capabilities. Failure (1): The agent guesses a customer case ID from conversation tone. Failure (2): The prompt hardcodes pricing that the app never updates. Which TWO specific corrections should the team apply?

  1. Require source comparison and abstention on missing evidence
  2. Enforce permitted methods at the service boundary
  3. Reject claims with no matching source evidence
  4. Move volatile facts into a governed retrieval source
  5. Require explicit identifiers and reject invented IDs

Explanation

Correct Answers

D:

Correct for failure (2). Static prompts are unsuitable as a changing source of commercial truth. It independently corrects the second defect.

E:

Correct for failure (1). Prompts can define required preconditions but tools must still enforce them. The proposed correction directly addresses that reported defect.

Incorrect Answers

A:

Grounding constrains the answer to the available text and provenance. The stated pair of failures instead requires “Move volatile facts into a governed retrieval source” and “Require explicit identifiers and reject invented IDs”; this option cannot substitute for both.

B:

Least-privilege tool methods reduce accidental external actions. It does not complete the two decisions required in this question.

C:

Prompt instructions must be paired with evidence verification for consequential facts. The stated pair of failures instead requires “Move volatile facts into a governed retrieval source” and “Require explicit identifiers and reject invented IDs”; this option cannot substitute for both.

Question 415

An AI operations team is testing two distinct capabilities. Failure (1): Page numbers are discarded before field extraction. Failure (2): The analyzer produces only a generic image caption. Which TWO specific corrections should the team apply?

  1. Represent repeated fields as arrays and validate counts
  2. Keep transcript anchors and validate extracted tasks
  3. Configure OCR-derived field extraction for the lot code
  4. Use a suitable video analyzer with supported source format
  5. Retain page IDs and reconcile source order

Explanation

Correct Answers

C:

Correct for failure (2). Visual information extraction can map pixels to structured identifiers. It independently corrects the second defect.

E:

Correct for failure (1). Page order and references are part of document provenance. The proposed correction directly addresses that reported defect.

Incorrect Answers

A:

Repeated records require structured lists with item-level associations. It does not complete the two decisions required in this question.

B:

Transcription and summarization are separate results with different validation. The stated pair of failures instead requires “Configure OCR-derived field extraction for the lot code” and “Retain page IDs and reconcile source order”; this option cannot substitute for both.

D:

Multimodal extraction draws on visual and spoken evidence over time. The stated pair of failures instead requires “Configure OCR-derived field extraction for the lot code” and “Retain page IDs and reconcile source order”; this option cannot substitute for both.

Question 416

An Azure AI Fundamentals solution has this requirement: A helpdesk wants to route requests into billing, connectivity, or account categories. Which implementation most directly addresses it?

  1. Use object detection when each product needs a location
  2. Use text classification with labeled intent examples
  3. Use machine translation with locale-specific quality review
  4. Use regression or forecasting appropriate to time-series data

Explanation

Correct Answer

B:

Classification maps text to one of several predefined categories. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Object detectors return classes and spatial boundaries. That is not sufficient here: Classification maps text to one of several predefined categories.

C:

Translation transfers meaning across languages rather than changing topic labels. That is not sufficient here: Classification maps text to one of several predefined categories.

D:

Forecasts predict future numeric quantities rather than discrete tags. That is not sufficient here: Classification maps text to one of several predefined categories.

Question 417

An acceptance tester needs to validate: Process the audio transcript and optionally key video scenes before summarizing. Which observed result proves the intended behavior?

  1. The timeline marks arrival frames with appropriate timestamps
  2. The summary preserves the topic sequence and supported key points
  3. The assistant produces a relevant answer with approved citations
  4. Known spikes score unusually relative to a validated baseline

Explanation

Correct Answer

B:

Meeting summaries require temporal and spoken source content. The expected acceptance evidence is: The summary preserves the topic sequence and supported key points. This observation directly tests the stated behavior.

Incorrect Answers

A:

Video understanding retains order and time, unlike single-image tagging. This observation does not establish the required result: The summary preserves the topic sequence and supported key points.

C:

Generation produces novel text, unlike choosing a fixed class label. This observation does not establish the required result: The summary preserves the topic sequence and supported key points.

D:

Anomaly methods identify deviations rather than assigning preset incident labels. This observation does not establish the required result: The summary preserves the topic sequence and supported key points.

Question 418

Troubleshooting report: Plain text without timestamps is fed to a caption player. Which correction is MOST directly targeted to this failure?

  1. Select the correct source/target language translation path
  2. Use the correct audio encoding and recognized input format
  3. Request time-aligned recognition details
  4. Select an appropriate audio-input model and upload the audio correctly

Explanation

Correct Answer

C:

This correction directly addresses the reported failure. Timecodes align text segments to audio playback. The targeted change is: Request time-aligned recognition details.

Incorrect Answers

A:

Translation requires semantic equivalence and language direction. In the stated case, troubleshooting report: Plain text without timestamps is fed to a caption player; “Select the correct source/target language translation path” does not target that failure.

B:

Speech recognition requires the service to decode audio correctly. In the stated case, troubleshooting report: Plain text without timestamps is fed to a caption player; “Use the correct audio encoding and recognized input format” does not target that failure.

D:

Model capabilities and accepted payload types differ across deployments. In the stated case, troubleshooting report: Plain text without timestamps is fed to a caption player; “Select an appropriate audio-input model and upload the audio correctly” does not target that failure.

Question 419

Troubleshooting report: The team adds repeated spaces to plain text expecting fixed-length silence. Which correction is MOST directly targeted to this failure?

  1. Provide valid SSML break elements to the synthesis service
  2. Require source-attributed action extraction
  3. Apply recognized punctuation support and validate the rendered transcript
  4. Ground the summary and penalize unsupported additions

Explanation

Correct Answer

A:

This correction directly addresses the reported failure. SSML controls aspects of prosody and timing for synthesized speech. The targeted change is: Provide valid SSML break elements to the synthesis service.

Incorrect Answers

B:

Transcription obtains words; summarization extracts decisions and tasks. In the stated case, troubleshooting report: The team adds repeated spaces to plain text expecting fixed-length silence; “Require source-attributed action extraction” does not target that failure.

C:

Speech text may require punctuation inference distinct from word recognition. In the stated case, troubleshooting report: The team adds repeated spaces to plain text expecting fixed-length silence; “Apply recognized punctuation support and validate the rendered transcript” does not target that failure.

D:

Summarization compresses written information rather than changing its facts. In the stated case, troubleshooting report: The team adds repeated spaces to plain text expecting fixed-length silence; “Ground the summary and penalize unsupported additions” does not target that failure.

Question 420

Troubleshooting report: The model gives answers without loading approved manuals. Which correction is MOST directly targeted to this failure?

  1. Use layout-aware structured invoice extraction
  2. Detect people and helmets and evaluate spatial associations
  3. Introduce a document classifier and route by type
  4. Add retrieval and grounded citation generation

Explanation

Correct Answer

D:

This correction directly addresses the reported failure. RAG retrieves current external evidence for a generated response. The targeted change is: Add retrieval and grounded citation generation.

Incorrect Answers

A:

Table extraction associates quantities, descriptions and prices by row. In the stated case, troubleshooting report: The model gives answers without loading approved manuals; “Use layout-aware structured invoice extraction” does not target that failure.

B:

Detection relates objects to positions for compliance checks. In the stated case, troubleshooting report: The model gives answers without loading approved manuals; “Detect people and helmets and evaluate spatial associations” does not target that failure.

C:

Classification selects the extraction process matching the document category. In the stated case, troubleshooting report: The model gives answers without loading approved manuals; “Introduce a document classifier and route by type” does not target that failure.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests