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 41

Question 400

An Azure AI Fundamentals solution has this requirement: A PDF contains a three-column price table and a computed total. Which implementation most directly addresses it?

  1. Use timestamped multi-segment video analysis for visible actions, with separate higher-frame-rate review when brief security events matter
  2. Preserve page identifiers and validate page sequence during extraction
  3. Preserve table row-column structure and validate numeric totals
  4. Check feature, model and regional availability before creating an analyzer

Explanation

Correct Answer

C:

Cell relationships are necessary to check item totals. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Content Understanding samples approximately one frame per second, so brief or single-frame actions can be missed despite valid timestamps. That is not sufficient here: Cell relationships are necessary to check item totals.

B:

Page order and references are part of document provenance. That is not sufficient here: Cell relationships are necessary to check item totals.

D:

Some modalities and features depend on resource and API version. That is not sufficient here: Cell relationships are necessary to check item totals.

Question 401

An acceptance tester needs to validate: Use a carefully constrained prompt and a lower sampling temperature where supported. Which observed result proves the intended behavior?

  1. The summary references events visible on frames and spoken in the transcript
  2. Repeated structured cases follow the expected schema more consistently
  3. The output returns evaluated sentiment classes rather than generated images
  4. A regression test records model version and key generation settings

Explanation

Correct Answer

B:

Lower randomness often improves consistency but does not guarantee correctness. The expected acceptance evidence is: Repeated structured cases follow the expected schema more consistently. This observation directly tests the stated behavior.

Incorrect Answers

A:

Different media channels provide complementary scene and dialogue evidence. This observation does not establish the required result: Repeated structured cases follow the expected schema more consistently.

C:

Different model tasks require compatible outputs and labeled evaluation data. This observation does not establish the required result: Repeated structured cases follow the expected schema more consistently.

D:

Reproducibility depends on more than the visible prompt text. This observation does not establish the required result: Repeated structured cases follow the expected schema more consistently.

Question 402

A solution team has selected this control: Preserve text spans and associated source regions during OCR. Which observation is the most direct acceptance evidence?

  1. Reviewers can inspect the handwritten comment's location
  2. Observed actions link to supported source intervals while unobserved short events are not claimed as detected
  3. The missing field is returned as null or flagged unknown
  4. The extracted steps map to the correct clips and source times

Explanation

Correct Answer

A:

Source-location evidence helps verify recognition on mixed-layout documents. The expected acceptance evidence is: Reviewers can inspect the handwritten comment's location. This observation directly tests the stated behavior.

Incorrect Answers

B:

Content Understanding samples approximately one frame per second, so brief or single-frame actions can be missed despite valid timestamps. This observation does not establish the required result: Reviewers can inspect the handwritten comment's location.

C:

Extraction should preserve absence and distinguish it from zero. This observation does not establish the required result: Reviewers can inspect the handwritten comment's location.

D:

Multimodal extraction draws on visual and spoken evidence over time. This observation does not establish the required result: Reviewers can inspect the handwritten comment's location.

Question 403

Troubleshooting report: An agent treats a page's embedded commands as system messages. Which correction is MOST directly targeted to this failure?

  1. Audit and correct all cross-region ancillary data flows
  2. Partition state by user and clear or secure session context
  3. Use layout-preserving extraction and test alternatives
  4. Separate retrieved data from tool and system instruction channels

Explanation

Correct Answer

D:

This correction directly addresses the reported failure. Lower-trust text is not a privileged command channel. The targeted change is: Separate retrieved data from tool and system instruction channels.

Incorrect Answers

A:

Privacy and compliance require controlling the full data path. In the stated case, troubleshooting report: An agent treats a page's embedded commands as system messages; “Audit and correct all cross-region ancillary data flows” does not target that failure.

B:

Privacy includes preventing cross-session memory disclosure. In the stated case, troubleshooting report: An agent treats a page's embedded commands as system messages; “Partition state by user and clear or secure session context” does not target that failure.

C:

Inclusion requires robust input handling beyond a single document template. In the stated case, troubleshooting report: An agent treats a page's embedded commands as system messages; “Use layout-preserving extraction and test alternatives” does not target that failure.

Question 404

An engineering review documents this requirement: An AI application processes client contracts from an unapproved geographic region. Select the most appropriate solution.

  1. Restrict processing, inference, logs and backups to authorized regions
  2. Represent uncertainty explicitly and refer to approved teaching sources
  3. Use grounded medical content, high-risk detection and appropriate expert escalation
  4. Include a verifiable source link and section reference alongside material policy answers

Explanation

Correct Answer

A:

Privacy and compliance require controlling the full data path. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Honesty requires acknowledging when a claim cannot be substantiated. That is not sufficient here: Privacy and compliance require controlling the full data path.

C:

Safety mechanisms should reduce unsupported high-impact recommendations. That is not sufficient here: Privacy and compliance require controlling the full data path.

D:

Transparency lets users inspect evidence instead of accepting fluent text. That is not sufficient here: Privacy and compliance require controlling the full data path.

Question 405

Troubleshooting report: The app calls a text classification API and expects a picture. Which correction is MOST directly targeted to this failure?

  1. Choose segmentation with validated pixel-level damage regions
  2. Choose a text-to-image generator instead of a text classification API
  3. Run unsupervised clustering and evaluate cluster characteristics
  4. Use key phrase extraction rather than regression

Explanation

Correct Answer

B:

This correction directly addresses the reported failure. Text-to-image generation creates new visual content from a description; classification only assigns labels to existing inputs. The targeted change is: Choose a text-to-image generator instead of a text classification API.

Incorrect Answers

A:

Pixel-level region masks localize building damage more precisely than an overall image class label. In the stated case, troubleshooting report: The app calls a text classification API and expects a picture; “Choose segmentation with validated pixel-level damage regions” does not target that failure.

C:

Clustering discovers groups based on feature similarity. In the stated case, troubleshooting report: The app calls a text classification API and expects a picture; “Run unsupervised clustering and evaluate cluster characteristics” does not target that failure.

D:

Key phrase extraction finds prominent topics without inventing new source material. In the stated case, troubleshooting report: The app calls a text classification API and expects a picture; “Use key phrase extraction rather than regression” does not target that failure.

Question 406

An Azure AI solution uses the following approved control: Use supported language identification before or during speech recognition. What observable result should pass QA?

  1. The result contains central complaint phrases without invented terms
  2. The sanitized SSML parses and renders the intended characters
  3. Silent input yields a repeat or timeout response rather than a denial
  4. Mixed test calls identify the correct language before transcription

Explanation

Correct Answer

D:

Automatic language detection chooses among an approved set of languages. The expected acceptance evidence is: Mixed test calls identify the correct language before transcription. This observation directly tests the stated behavior.

Incorrect Answers

A:

Key phrases highlight important concepts expressed in the original text. This observation does not establish the required result: Mixed test calls identify the correct language before transcription.

B:

SSML is XML and malformed markup breaks the synthesis request. This observation does not establish the required result: Mixed test calls identify the correct language before transcription.

C:

An empty recognition result is not semantically equivalent to the word no. This observation does not establish the required result: Mixed test calls identify the correct language before transcription.

Question 407

An Azure AI solution uses the following approved control: Build a supported messages collection with roles and the user's content. What observable result should pass QA?

  1. The service processes role-tagged messages with the expected application behavior
  2. The target model deploys in the approved region and meets residency controls
  3. A generated claim resolves to an authorized supporting source
  4. The same task passes with reduced token use and no quality loss

Explanation

Correct Answer

A:

Chat APIs use structured messages rather than concatenating untyped command strings. The expected acceptance evidence is: The service processes role-tagged messages with the expected application behavior. This observation directly tests the stated behavior.

Incorrect Answers

B:

Not every model or feature is available in every location. This observation does not establish the required result: The service processes role-tagged messages with the expected application behavior.

C:

Prompt instructions must be paired with evidence verification for consequential facts. This observation does not establish the required result: The service processes role-tagged messages with the expected application behavior.

D:

Long irrelevant context consumes tokens and may increase latency. This observation does not establish the required result: The service processes role-tagged messages with the expected application behavior.

Question 408

The approved design action is: Use text embeddings with nearest-neighbor search and business filters. Which test result would BEST establish that it worked?

  1. Parsed fields match labeled form values and missing fields remain explicit
  2. Names in a labeled sample are detected and redacted
  3. The model returns the sign category and any required location
  4. Suggested products are relevant and meet inventory/authorization rules

Explanation

Correct Answer

D:

Semantic vectors capture related concepts while filters enforce eligibility. The expected acceptance evidence is: Suggested products are relevant and meet inventory/authorization rules. This observation directly tests the stated behavior.

Incorrect Answers

A:

Document extraction maps unstructured layouts into structured values. This observation does not establish the required result: Suggested products are relevant and meet inventory/authorization rules.

B:

Entity extraction locates sensitive spans for privacy handling. This observation does not establish the required result: Suggested products are relevant and meet inventory/authorization rules.

C:

The task chooses a label, with localization needed only when multiple signs appear. This observation does not establish the required result: Suggested products are relevant and meet inventory/authorization rules.

Question 409

An AI operations team is testing two distinct capabilities. Failure (1): The service returns paragraphs instead of requested tag labels. Failure (2): The app provides one global category for several separate objects. Which TWO specific corrections should the team apply?

  1. Request tags and parse the tag collection
  2. Use the documented coordinate units and image dimensions
  3. Improve the product mask and restrict the editable region
  4. Switch to object detection for localized results
  5. Review multiple timestamped frames and flag rapid or unobserved steps instead of inferring them

Explanation

Correct Answers

A:

Correct for failure (1). Tags summarize recognized concepts as labels. The proposed correction directly addresses that reported defect.

D:

Correct for failure (2). Classification predicts image-level categories while detection provides positions. It independently corrects the second defect.

Incorrect Answers

B:

Localization metadata is tied to pixel regions, not just labels. The stated pair of failures instead requires “Request tags and parse the tag collection” and “Switch to object detection for localized results”; this option cannot substitute for both.

C:

This action would address another problem: the edit changes the product logo while replacing the background. A mask defines which image area may change. It does not resolve either specified failure.

E:

A single still cannot convey action order; sampled video frames may also omit fleeting intermediate steps. The stated pair of failures instead requires “Request tags and parse the tag collection” and “Switch to object detection for localized results”; this option cannot substitute for both.

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.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests