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 77

Question 760

Troubleshooting report: The pipeline performs only speech recognition and produces no audio reply. Which correction is MOST directly targeted to this failure?

  1. Add realistic noise tests and a clarification fallback
  2. Add an authorized synthesis stage for the answer
  3. Supply an authorized identity and matching resource configuration
  4. Add a no-speech branch and clarification option

Explanation

Correct Answer

B:

This correction directly addresses the reported failure. Input transcription and output synthesis are distinct steps. The targeted change is: Add an authorized synthesis stage for the answer.

Incorrect Answers

A:

Signal quality directly affects recognizer error rates. In the stated case, troubleshooting report: The pipeline performs only speech recognition and produces no audio reply; “Add realistic noise tests and a clarification fallback” does not target that failure.

C:

Audio capture success does not establish remote service authentication. In the stated case, troubleshooting report: The pipeline performs only speech recognition and produces no audio reply; “Supply an authorized identity and matching resource configuration” does not target that failure.

D:

An empty recognition result is not semantically equivalent to the word no. In the stated case, troubleshooting report: The pipeline performs only speech recognition and produces no audio reply; “Add a no-speech branch and clarification option” does not target that failure.

Question 761

Troubleshooting report: The system merges checkbox states into surrounding text. Which correction is MOST directly targeted to this failure?

  1. Use document layout extraction with typed checkbox outputs
  2. Replace the TTS stage with recognition/transcription
  3. Use layout-aware structured invoice extraction
  4. Define extraction fields and validate values

Explanation

Correct Answer

A:

This correction directly addresses the reported failure. Different form elements require more than ordinary running-text OCR. The targeted change is: Use document layout extraction with typed checkbox outputs.

Incorrect Answers

B:

Speech recognition converts speech into words suitable for search. In the stated case, troubleshooting report: The system merges checkbox states into surrounding text; “Replace the TTS stage with recognition/transcription” does not target that failure.

C:

Table extraction associates quantities, descriptions and prices by row. In the stated case, troubleshooting report: The system merges checkbox states into surrounding text; “Use layout-aware structured invoice extraction” does not target that failure.

D:

Document extraction maps unstructured layouts into structured values. In the stated case, troubleshooting report: The system merges checkbox states into surrounding text; “Define extraction fields and validate values” does not target that failure.

Question 762

The approved design action is: Validate file type, size and image dimensions before the model call. Which test result would BEST establish that it worked?

  1. The returned string matches the labeled serial number
  2. A returned serial number is stored as OCR text not a generic image tag
  3. A drawn overlay aligns each box with the corresponding person
  4. A correctly converted test image processes successfully

Explanation

Correct Answer

D:

Vision endpoints impose supported input formats and limits. The expected acceptance evidence is: A correctly converted test image processes successfully. This observation directly tests the stated behavior.

Incorrect Answers

A:

OCR extracts characters visible in pixels. This observation does not establish the required result: A correctly converted test image processes successfully.

B:

Tags describe concepts while OCR reports character content. This observation does not establish the required result: A correctly converted test image processes successfully.

C:

Localization metadata is tied to pixel regions, not just labels. This observation does not establish the required result: A correctly converted test image processes successfully.

Question 763

A solution team has selected this control: Keep retrieved documents in lower-trust context and filter tool requests independently. Which observation is the most direct acceptance evidence?

  1. An injected page fails to invoke an unauthorized tool
  2. Browser traffic cannot access the protected model credential
  3. The second tenant's test passes with its own authorized identity
  4. A generated claim resolves to an authorized supporting source

Explanation

Correct Answer

A:

External content cannot replace privileged system behavior. The expected acceptance evidence is: An injected page fails to invoke an unauthorized tool. This observation directly tests the stated behavior.

Incorrect Answers

B:

Backend controls protect credentials, quotas and data policy. This observation does not establish the required result: An injected page fails to invoke an unauthorized tool.

C:

Copying code does not grant access to another tenant's resources. This observation does not establish the required result: An injected page fails to invoke an unauthorized tool.

D:

Prompt instructions must be paired with evidence verification for consequential facts. This observation does not establish the required result: An injected page fails to invoke an unauthorized tool.

Question 764

An acceptance tester needs to validate: Define a custom analyzer for consistent target field meanings across templates. Which observed result proves the intended behavior?

  1. The checked option is returned accurately as true or selected
  2. Several labeled forms populate the same typed output fields
  3. Each date field links to the correct labeled region
  4. Speaker assignments align to audible turns with flagged overlaps

Explanation

Correct Answer

B:

A schema-driven analyzer identifies business fields independent of fixed coordinates. The expected acceptance evidence is: Several labeled forms populate the same typed output fields. This observation directly tests the stated behavior.

Incorrect Answers

A:

Checkbox state is a structured signal rather than arbitrary prose. This observation does not establish the required result: Several labeled forms populate the same typed output fields.

C:

Typed extraction avoids mixing semantically different dates. This observation does not establish the required result: Several labeled forms populate the same typed output fields.

D:

Conversation extraction needs more than a single flattened text stream. This observation does not establish the required result: Several labeled forms populate the same typed output fields.

Question 765

An AI operations team is testing two distinct capabilities. Failure (1): The avatar uses an employee's identity without disclosure. Failure (2): The audit trail contains only a free-form answer string. Which TWO specific corrections should the team apply?

  1. Replace impersonating presentation with explicit synthetic-agent notice
  2. Add an accessible name and keyboard activation to the control
  3. Execute and audit the supported deletion workflow
  4. Put an explicit confirmation gate before high-impact actions
  5. Archive the actual decision and versioned provenance with privacy minimization

Explanation

Correct Answers

A:

Correct for failure (1). Transparency prevents users from falsely assuming they are speaking with a human. The proposed correction directly addresses that reported defect.

E:

Correct for failure (2). Auditability requires the original decision evidence and version identifiers; regenerating model text is not guaranteed to be identical. It independently corrects the second defect.

Incorrect Answers

B:

Inclusive design permits nonvisual navigation of the same task path. The stated pair of failures instead requires “Replace impersonating presentation with explicit synthetic-agent notice” and “Archive the actual decision and versioned provenance with privacy minimization”; this option cannot substitute for both.

C:

Accountability requires evidence of actual data lifecycle actions. It does not complete the two decisions required in this question.

D:

Reliable assistance distinguishes missing intent from authorized action. It does not complete the two decisions required in this question.

Question 766

An AI operations team is testing two distinct capabilities. Failure (1): The acceptance report is only a generated paragraph saying all rules passed. Failure (2): All policy checks run only on the initial user message. Which TWO specific corrections should the team apply?

  1. Raise high-risk routing and restrict unsupported actionable advice
  2. Permit calibrated caveats and qualified review paths
  3. Add independent risk review and verifiable test artifacts
  4. Rebalance safety tests and adjust erroneous dialect associations
  5. Add post-retrieval and pre-action policy enforcement

Explanation

Correct Answers

C:

Correct for failure (1). Self-attestation is not independent proof of regulatory controls. The proposed correction directly addresses that reported defect.

E:

Correct for failure (2). Threats can enter later stages even when the original user request was harmless. It independently corrects the second defect.

Incorrect Answers

A:

Safety mechanisms should reduce unsupported high-impact recommendations. It does not complete the two decisions required in this question.

B:

Transparency reduces mistaken reliance on unsupported professional guarantees. It does not complete the two decisions required in this question.

D:

Fairness requires discriminating actual harm from linguistic variety. It does not complete the two decisions required in this question.

Question 767

An engineering review documents this requirement: An assistant must plan a multi-step trip reservation but should not spend money automatically. Select the most appropriate solution.

  1. Use image editing or segmentation with object-mask preservation
  2. Use supervised classification with labeled maintenance outcomes
  3. Use invoice-oriented structured extraction with table row preservation
  4. Use an agent orchestration plan with separate approval before booking

Explanation

Correct Answer

D:

An agent may propose actions but payment must remain permission-controlled. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Editing changes selected pixels while leaving product regions intact. That is not sufficient here: An agent may propose actions but payment must remain permission-controlled.

B:

A failure probability or class is a discrete decision under a defined horizon. That is not sufficient here: An agent may propose actions but payment must remain permission-controlled.

C:

Table extraction associates quantities, descriptions and prices by row. That is not sufficient here: An agent may propose actions but payment must remain permission-controlled.

Question 768

An AI operations team is testing two distinct capabilities. Failure (1): The system merely replies that the data was forgotten. Failure (2): The test set has no metadata about device types. Which TWO specific corrections should the team apply?

  1. Reconcile the provider transaction ID and actual status before confirmation
  2. Replace impersonating presentation with explicit synthetic-agent notice
  3. Label and evaluate representative scans by acquisition equipment
  4. Apply data minimization, retention and permission controls
  5. Execute and audit the supported deletion workflow

Explanation

Correct Answers

C:

Correct for failure (2). Reliability requires detecting distribution shifts hidden in aggregate accuracy. It independently corrects the second defect.

E:

Correct for failure (1). Accountability requires evidence of actual data lifecycle actions. The proposed correction directly addresses that reported defect.

Incorrect Answers

A:

Generated text is not evidence of external side effects. The stated pair of failures instead requires “Label and evaluate representative scans by acquisition equipment” and “Execute and audit the supported deletion workflow”; this option cannot substitute for both.

B:

Transparency prevents users from falsely assuming they are speaking with a human. The stated pair of failures instead requires “Label and evaluate representative scans by acquisition equipment” and “Execute and audit the supported deletion workflow”; this option cannot substitute for both.

D:

Privacy requires meaningful notice and bounded handling of personal data. The stated pair of failures instead requires “Label and evaluate representative scans by acquisition equipment” and “Execute and audit the supported deletion workflow”; this option cannot substitute for both.

Question 769

An acceptance tester needs to validate: Stop active recording and apply the documented microphone permission state. Which observed result proves the intended behavior?

  1. The transcript marks each turn with a stable speaker label
  2. The converted test recording produces a faithful transcript
  3. The sanitized SSML parses and renders the intended characters
  4. Mic capture ceases and the app no longer receives audio frames

Explanation

Correct Answer

D:

User control of capture must be enforced at the device or application layer. The expected acceptance evidence is: Mic capture ceases and the app no longer receives audio frames. This observation directly tests the stated behavior.

Incorrect Answers

A:

Speaker diarization separates turns and identities at a conversation level. This observation does not establish the required result: Mic capture ceases and the app no longer receives audio frames.

B:

Speech recognition requires the service to decode audio correctly. This observation does not establish the required result: Mic capture ceases and the app no longer receives audio frames.

C:

SSML is XML and malformed markup breaks the synthesis request. This observation does not establish the required result: Mic capture ceases and the app no longer receives audio frames.

Question 770

An engineering review documents this requirement: A developer needs to create a chatbot with a persistent corporate system instruction. Select the most appropriate solution.

  1. Archive the actual evaluated response together with the prompt, model deployment, input and generation settings
  2. Store the approved instruction as versioned application configuration
  3. Add an approved knowledge/retrieval tool with clear description and access limits
  4. Limit retrieval and chat context to the authenticated case scope

Explanation

Correct Answer

B:

Versioned system messages provide a controlled behavioral baseline. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Recorded settings support investigation, but generative models can still vary on an apparently identical rerun. That is not sufficient here: Versioned system messages provide a controlled behavioral baseline.

C:

Tool descriptions guide selection, while backend rules enforce allowed sources. That is not sufficient here: Versioned system messages provide a controlled behavioral baseline.

D:

Source scoping enforces privacy and also improves answer relevance. That is not sufficient here: Versioned system messages provide a controlled behavioral baseline.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests