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 83

Question 820

Troubleshooting report: The output accidentally includes content from an unrelated uploaded image. Which correction is MOST directly targeted to this failure?

  1. Use image input and OCR or vision grounding
  2. Route to the text-to-image mode with intended inputs only
  3. Request the needed visual features explicitly
  4. Gate distribution on the image moderation result

Explanation

Correct Answer

B:

This correction directly addresses the reported failure. Generation from text is distinct from image editing. The targeted change is: Route to the text-to-image mode with intended inputs only.

Incorrect Answers

A:

Visual evidence is needed to read sign content. In the stated case, troubleshooting report: The output accidentally includes content from an unrelated uploaded image; “Use image input and OCR or vision grounding” does not target that failure.

C:

Different visual outputs describe text, scene classes and detected people. In the stated case, troubleshooting report: The output accidentally includes content from an unrelated uploaded image; “Request the needed visual features explicitly” does not target that failure.

D:

Generative media needs policy controls in addition to prompt instructions. In the stated case, troubleshooting report: The output accidentally includes content from an unrelated uploaded image; “Gate distribution on the image moderation result” does not target that failure.

Question 821

Troubleshooting report: The caption contains unsupported details about unseen objects. Which correction is MOST directly targeted to this failure?

  1. Add confidence gating and supporting evidence
  2. Ground the alt text in the actual picture
  3. Request the needed visual features explicitly
  4. Apply a visual acceptance review and keep the strongest matching variants

Explanation

Correct Answer

B:

This correction directly addresses the reported failure. Accessible captions should communicate image purpose without invented details. The targeted change is: Ground the alt text in the actual picture.

Incorrect Answers

A:

Generated visual interpretations can be wrong even when phrased fluently. In the stated case, troubleshooting report: The caption contains unsupported details about unseen objects; “Add confidence gating and supporting evidence” does not target that failure.

C:

Different visual outputs describe text, scene classes and detected people. In the stated case, troubleshooting report: The caption contains unsupported details about unseen objects; “Request the needed visual features explicitly” does not target that failure.

D:

Image-generation variability can be useful but does not guarantee suitability. In the stated case, troubleshooting report: The caption contains unsupported details about unseen objects; “Apply a visual acceptance review and keep the strongest matching variants” does not target that failure.

Question 822

Troubleshooting report: The API receives a single malformed string instead of required message structure. Which correction is MOST directly targeted to this failure?

  1. Benchmark candidate deployments and choose by requirements
  2. Construct messages in the documented SDK request format
  3. Tighten the output contract and apply post-response validation
  4. Keep publication behind a separate approved deployment step

Explanation

Correct Answer

B:

This correction directly addresses the reported failure. Chat APIs use structured messages rather than concatenating untyped command strings. The targeted change is: Construct messages in the documented SDK request format.

Incorrect Answers

A:

Lower-cost deployment is acceptable only if accuracy and safety floors remain met. In the stated case, troubleshooting report: The API receives a single malformed string instead of required message structure; “Benchmark candidate deployments and choose by requirements” does not target that failure.

C:

Explicit examples can improve adherence to bounded output conventions. In the stated case, troubleshooting report: The API receives a single malformed string instead of required message structure; “Tighten the output contract and apply post-response validation” does not target that failure.

D:

A draft or tested agent is not automatically a publicly released service. In the stated case, troubleshooting report: The API receives a single malformed string instead of required message structure; “Keep publication behind a separate approved deployment step” does not target that failure.

Question 823

An Azure AI Fundamentals solution has this requirement: A manufacturer needs to distinguish two similar product defects in images. Which implementation most directly addresses it?

  1. Use multiple timestamped representative video frames to describe visible steps, without inventing unseen rapid events
  2. Preserve region coordinates with extracted text or field evidence
  3. Evaluate a visual classification or detection model on labeled examples of both defects
  4. Use an image-generation-capable model with an appropriate prompt

Explanation

Correct Answer

C:

Fine-grained visual classes require representative discriminating images. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

A single still cannot convey action order; sampled video frames may also omit fleeting intermediate steps. That is not sufficient here: Fine-grained visual classes require representative discriminating images.

B:

Spatial provenance supports later verification of recognition results. That is not sufficient here: Fine-grained visual classes require representative discriminating images.

D:

Image generation creates new pixels rather than classifying an existing photo. That is not sufficient here: Fine-grained visual classes require representative discriminating images.

Question 824

An Azure-based AI team must satisfy this objective: An agent ingests audio and video files but users only need a short source-linked answer. Which design choice is best?

  1. Enable a supported barcode extraction or recognition feature
  2. Choose media analyzers with transcript and scene evidence tied to the retrieved answer
  3. Use supported speaker/timing evidence and keep uncertain attribution explicit
  4. Recompute arithmetic and compare totals with reported invoice fields

Explanation

Correct Answer

B:

Multimodal grounding should preserve how the answer maps to source media. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

A barcode is a structured visual symbol not ordinary OCR text. That is not sufficient here: Multimodal grounding should preserve how the answer maps to source media.

C:

Conversation extraction needs more than a single flattened text stream. That is not sufficient here: Multimodal grounding should preserve how the answer maps to source media.

D:

Independent business rules catch plausible extraction errors. That is not sufficient here: Multimodal grounding should preserve how the answer maps to source media.

Question 825

A solution team has selected this control: Prefer the GA API version and features unless preview use is explicitly approved. Which observation is the most direct acceptance evidence?

  1. The deployed analyzer uses a documented production-capable version
  2. Observed actions link to supported source intervals while unobserved short events are not claimed as detected
  3. The resulting record maps each value to its correct page
  4. The result reports both candidate values and cites their locations

Explanation

Correct Answer

A:

Preview-only features may have different availability and contracts. The expected acceptance evidence is: The deployed analyzer uses a documented production-capable version. 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: The deployed analyzer uses a documented production-capable version.

C:

Page order and references are part of document provenance. This observation does not establish the required result: The deployed analyzer uses a documented production-capable version.

D:

A model should not silently choose among competing contractual evidence. This observation does not establish the required result: The deployed analyzer uses a documented production-capable version.

Question 826

An Azure-based AI team must satisfy this objective: A model evaluation declares full compliance without recording who reviewed exceptions. Which design choice is best?

  1. Maintain a governed feature inventory and trace decisions to policy
  2. Exclude protected proxies from eligibility logic and test the remaining permitted factors
  3. Require governance evidence with reviewers, test outcomes, and documented exceptions
  4. Treat retrieved content as untrusted evidence and prevent it from changing tool authority

Explanation

Correct Answer

C:

Self-attestation is not independent proof of regulatory controls. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Accountability requires interpretable reasons for consequential decisions. That is not sufficient here: Self-attestation is not independent proof of regulatory controls.

B:

Legitimate credit criteria cannot be replaced by sensitive identity traits. That is not sufficient here: Self-attestation is not independent proof of regulatory controls.

D:

Lower-trust text is not a privileged command channel. That is not sufficient here: Self-attestation is not independent proof of regulatory controls.

Question 827

An Azure-based AI team must satisfy this objective: A production assistant needs to handle messages with different system and user roles. Which design choice is best?

  1. Use a supervised classification model with representative labeled cases
  2. Place durable application behavior constraints in supported high-priority instructions
  3. Choose a compatible multimodal model and provide supported text/image inputs
  4. Choose a text-classification or sentiment-analysis capability rather than an image generator

Explanation

Correct Answer

B:

Instruction roles help distinguish application policy from user task text. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Classification predicts discrete labels rather than a continuous numeric target. That is not sufficient here: Instruction roles help distinguish application policy from user task text.

C:

Multimodal models can reason over more than one supported input modality. That is not sufficient here: Instruction roles help distinguish application policy from user task text.

D:

Different model tasks require compatible outputs and labeled evaluation data. That is not sufficient here: Instruction roles help distinguish application policy from user task text.

Question 828

An Azure AI solution uses the following approved control: Use a translation service or supported multilingual generation with evaluation. What observable result should pass QA?

  1. The translated request preserves numbers and intent
  2. The response contains both an attitude label and typed entity spans
  3. The listener hears equivalent French instructions
  4. Early spoken words are present in a captured test transcript

Explanation

Correct Answer

A:

Translation requires semantic equivalence and language direction. The expected acceptance evidence is: The translated request preserves numbers and intent. This observation directly tests the stated behavior.

Incorrect Answers

B:

Sentiment polarity and named entity spans answer different questions. This observation does not establish the required result: The translated request preserves numbers and intent.

C:

Cross-language audio output requires both language transfer and synthesis. This observation does not establish the required result: The translated request preserves numbers and intent.

D:

Real-time audio pipelines need coordinated capture and recognition start. This observation does not establish the required result: The translated request preserves numbers and intent.

Question 829

A solution team has selected this control: Use layout-aware extraction and selection-mark analysis. Which observation is the most direct acceptance evidence?

  1. Read requests succeed while unauthorized writes are blocked
  2. Extracted checks and notes match visible form locations
  3. The minutes correctly associate tasks with speakers and source timestamps
  4. An article yields correctly labeled organizations and products

Explanation

Correct Answer

B:

Different form elements require more than ordinary running-text OCR. The expected acceptance evidence is: Extracted checks and notes match visible form locations. This observation directly tests the stated behavior.

Incorrect Answers

A:

Agentic workflows combine information access with controlled actions. This observation does not establish the required result: Extracted checks and notes match visible form locations.

C:

Attribution distinguishes who committed to each action. This observation does not establish the required result: Extracted checks and notes match visible form locations.

D:

NER locates and categorizes named entities in text. This observation does not establish the required result: Extracted checks and notes match visible form locations.

Question 830

An Azure AI Fundamentals solution has this requirement: A support app uses a tool that should only list orders. Which implementation most directly addresses it?

  1. Expose a read-only tool contract rather than a broad administrative API
  2. Preserve the higher-priority application constraints in the message hierarchy
  3. Use operation IDs and read-back receipts before replaying external side effects
  4. Honor service throttling and use bounded exponential backoff

Explanation

Correct Answer

A:

Least-privilege tool methods reduce accidental external actions. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Developer and system constraints govern allowed operations over user requests. That is not sufficient here: Least-privilege tool methods reduce accidental external actions.

C:

A timeout does not prove that the original mutation failed. That is not sufficient here: Least-privilege tool methods reduce accidental external actions.

D:

Rate limits protect the service and uncontrolled retries amplify overload. That is not sufficient here: Least-privilege tool methods reduce accidental external actions.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests