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 58

Question 570

An Azure AI solution uses the following approved control: Use validated data retrieval and source-backed numerical responses. What observable result should pass QA?

  1. A rejected test case can be reviewed with its supporting evidence
  2. Eligible memory entries disappear from active stores after the verified request
  3. A first-use test displays clear AI identification and human contact option
  4. The returned metric reconciles to the approved source table

Explanation

Correct Answer

D:

A numeric-looking answer is not proof that a measure was computed. The expected acceptance evidence is: The returned metric reconciles to the approved source table. This observation directly tests the stated behavior.

Incorrect Answers

A:

Accountability includes traceable responsibility and appropriate contestability. This observation does not establish the required result: The returned metric reconciles to the approved source table.

B:

Accountability requires evidence of actual data lifecycle actions. This observation does not establish the required result: The returned metric reconciles to the approved source table.

C:

Transparency helps users assess the scope and limitations of the assistance. This observation does not establish the required result: The returned metric reconciles to the approved source table.

Question 571

An Azure AI Fundamentals solution has this requirement: A customer support chatbot must show where a policy claim came from. Which implementation most directly addresses it?

  1. Include a verifiable source link and section reference alongside material policy answers
  2. Set explicit uncertainty thresholds and route high-risk decisions to qualified review
  3. Measure recognition errors across representative accent and speaking groups
  4. Restrict retrieval to effective approved documents and expose source dates

Explanation

Correct Answer

A:

Transparency lets users inspect evidence instead of accepting fluent text. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

Safety-critical recommendations require verified constraints outside open-ended generation. That is not sufficient here: Transparency lets users inspect evidence instead of accepting fluent text.

C:

Inclusive voice systems require performance evaluation beyond a single accent. That is not sufficient here: Transparency lets users inspect evidence instead of accepting fluent text.

D:

Truthfulness requires evidence that corresponds to the operative version. That is not sufficient here: Transparency lets users inspect evidence instead of accepting fluent text.

Question 572

An engineering review documents this requirement: A customer support desk receives spoken customer questions and needs searchable text. Select the most appropriate solution.

  1. Use speech-to-text transcription before indexing call content
  2. Use conversational generative AI with grounded retrieval and scoped dialogue state
  3. Use semantic document indexing and retrieval rather than only keyword equality
  4. Use object detection for helmet presence and worker locations

Explanation

Correct Answer

A:

Speech recognition converts speech into words suitable for search. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

B:

The assistant should incorporate context while remaining connected to reliable sources. That is not sufficient here: Speech recognition converts speech into words suitable for search.

C:

Semantic search can retrieve conceptually related passages. That is not sufficient here: Speech recognition converts speech into words suitable for search.

D:

Detection relates objects to positions for compliance checks. That is not sufficient here: Speech recognition converts speech into words suitable for search.

Question 573

An Azure AI Fundamentals solution has this requirement: A model generates textual descriptions but cannot produce an image file. Which implementation most directly addresses it?

  1. Adjust supported sampling parameters while preserving safety and quality tests
  2. Use a carefully constrained prompt and a lower sampling temperature where supported
  3. Choose a compatible multimodal model and provide supported text/image inputs
  4. Select an image-generation-capable model and a supported image endpoint

Explanation

Correct Answer

D:

Text generation and image generation are distinct model output capabilities. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

More varied token sampling can increase creative alternatives. That is not sufficient here: Text generation and image generation are distinct model output capabilities.

B:

Lower randomness often improves consistency but does not guarantee correctness. That is not sufficient here: Text generation and image generation are distinct model output capabilities.

C:

Multimodal models can reason over more than one supported input modality. That is not sufficient here: Text generation and image generation are distinct model output capabilities.

Question 574

An Azure AI solution uses the following approved control: Use tenant-specific identity, endpoint configuration and permissions. What observable result should pass QA?

  1. The second tenant's test passes with its own authorized identity
  2. The service processes role-tagged messages with the expected application behavior
  3. A smoke test reaches the intended new deployment and parses the output
  4. Browser traffic cannot access the protected model credential

Explanation

Correct Answer

A:

Copying code does not grant access to another tenant's resources. The expected acceptance evidence is: The second tenant's test passes with its own authorized identity. This observation directly tests the stated behavior.

Incorrect Answers

B:

Chat APIs use structured messages rather than concatenating untyped command strings. This observation does not establish the required result: The second tenant's test passes with its own authorized identity.

C:

A model family label is not necessarily the concrete deployed endpoint. This observation does not establish the required result: The second tenant's test passes with its own authorized identity.

D:

Backend controls protect credentials, quotas and data policy. This observation does not establish the required result: The second tenant's test passes with its own authorized identity.

Question 575

An AI operations team is testing two distinct capabilities. Failure (1): The developer interprets an entity-confidence score as customer sentiment. Failure (2): The service returns only an overall positive attitude. Which TWO specific corrections should the team apply?

  1. Keep operation outputs separate and map each to its intended field
  2. Load the deployed speech resource's region and endpoint
  3. Close the capture stream and update permission controls
  4. Use entity extraction instead of sentiment classification
  5. Use explicit language confirmation or a support fallback

Explanation

Correct Answers

A:

Correct for failure (1). Sentiment polarity and named entity spans answer different questions. The proposed correction directly addresses that reported defect.

D:

Correct for failure (2). NER identifies named spans rather than summarizing the email. It independently corrects the second defect.

Incorrect Answers

B:

Region or endpoint mismatch prevents requests reaching the correct resource. The stated pair of failures instead requires “Keep operation outputs separate and map each to its intended field” and “Use entity extraction instead of sentiment classification”; this option cannot substitute for both.

C:

User control of capture must be enforced at the device or application layer. The stated pair of failures instead requires “Keep operation outputs separate and map each to its intended field” and “Use entity extraction instead of sentiment classification”; this option cannot substitute for both.

E:

Unrecognized languages must not be silently interpreted as a known one. The stated pair of failures instead requires “Keep operation outputs separate and map each to its intended field” and “Use entity extraction instead of sentiment classification”; this option cannot substitute for both.

Question 576

An Azure AI solution uses the following approved control: Use an agent-capable orchestration flow with bounded tool definitions. What observable result should pass QA?

  1. A held-out set receives the correct whole-image classes and empirically calibrated class probabilities
  2. The execution trace shows a permitted tool called with validated inputs
  3. The same facts are consistently presented in the required style
  4. A permitted larger output budget allows the complete response

Explanation

Correct Answer

B:

Agent workflows can reason about tool selection but tools still need authorization. The expected acceptance evidence is: The execution trace shows a permitted tool called with validated inputs. This observation directly tests the stated behavior.

Incorrect Answers

A:

Classification assigns image-wide categories rather than bounding-box coordinates. This observation does not establish the required result: The execution trace shows a permitted tool called with validated inputs.

C:

Clear prompting can control tone without model retraining. This observation does not establish the required result: The execution trace shows a permitted tool called with validated inputs.

D:

Generation stops when an output budget or stop condition is reached. This observation does not establish the required result: The execution trace shows a permitted tool called with validated inputs.

Question 577

An Azure AI Fundamentals solution has this requirement: A team wants to inspect how an agent reached its final answer. Which implementation most directly addresses it?

  1. Archive the actual evaluated response together with the prompt, model deployment, input and generation settings
  2. Capture allowed tool calls, response statuses and source references in traces
  3. Run adversarial model evaluations and verify that configured content-safety guardrails enforce the stated high-severity block criteria
  4. Keep retrieved documents in lower-trust context and filter tool requests independently

Explanation

Correct Answer

B:

Operational observability aids debugging without exposing hidden sensitive prompt data. 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: Operational observability aids debugging without exposing hidden sensitive prompt data.

C:

Adversarial testing must measure actual configured policy enforcement rather than treating a prompt instruction as a guaranteed safety boundary. That is not sufficient here: Operational observability aids debugging without exposing hidden sensitive prompt data.

D:

External content cannot replace privileged system behavior. That is not sufficient here: Operational observability aids debugging without exposing hidden sensitive prompt data.

Question 578

The approved design action is: Apply retention and deletion controls subject to recorded legal exceptions. Which test result would BEST establish that it worked?

  1. The transfer remains pending until delegated approval is recorded
  2. A residency map confirms every protected copy remains in allowed locations
  3. Eligible memory entries disappear from active stores after the verified request
  4. A deaf tester receives the complete equivalent content as text

Explanation

Correct Answer

C:

Accountability requires evidence of actual data lifecycle actions. The expected acceptance evidence is: Eligible memory entries disappear from active stores after the verified request. This observation directly tests the stated behavior.

Incorrect Answers

A:

Accountable high-impact operations need verifiable authorized signatures. This observation does not establish the required result: Eligible memory entries disappear from active stores after the verified request.

B:

Privacy and compliance require controlling the full data path. This observation does not establish the required result: Eligible memory entries disappear from active stores after the verified request.

D:

Accessible information must not depend on hearing audio. This observation does not establish the required result: Eligible memory entries disappear from active stores after the verified request.

Question 579

Troubleshooting report: A document-only analyzer is called with a video input. Which correction is MOST directly targeted to this failure?

  1. Use a suitable video analyzer with supported source format
  2. Poll the documented operation status and handle failures
  3. Build citations from stored IDs and source spans
  4. Use barcode decoding and validate content

Explanation

Correct Answer

A:

This correction directly addresses the reported failure. Multimodal extraction draws on visual and spoken evidence over time. The targeted change is: Use a suitable video analyzer with supported source format.

Incorrect Answers

B:

Long-running media analysis may be asynchronous. In the stated case, troubleshooting report: A document-only analyzer is called with a video input; “Poll the documented operation status and handle failures” does not target that failure.

C:

Verifiable provenance differentiates a cited answer from guesswork. In the stated case, troubleshooting report: A document-only analyzer is called with a video input; “Build citations from stored IDs and source spans” does not target that failure.

D:

A barcode is a structured visual symbol not ordinary OCR text. In the stated case, troubleshooting report: A document-only analyzer is called with a video input; “Use barcode decoding and validate content” does not target that failure.

Question 580

An Azure AI Fundamentals solution has this requirement: A product Q&A assistant needs frequently revised catalog facts. Which implementation most directly addresses it?

  1. Use text-to-speech synthesis with a supported voice
  2. Use clustering or another appropriate unsupervised technique
  3. Use retrieval-augmented generation from current approved catalog sources
  4. Use a held-out evaluation and investigate overfitting or data leakage

Explanation

Correct Answer

C:

External retrieval updates factual context without retraining model weights. This implementation addresses the precise goal rather than an adjacent AI task.

Incorrect Answers

A:

Speech synthesis generates audio from text rather than transcribing audio. That is not sufficient here: External retrieval updates factual context without retraining model weights.

B:

Unsupervised learning finds structure without supplied target labels. That is not sufficient here: External retrieval updates factual context without retraining model weights.

D:

A model can memorize training examples without generalizing. That is not sufficient here: External retrieval updates factual context without retraining model weights.

Microsoft Azure AI Fundamentals (AI-901) Practice Tests