Microsoft AB-731 Practice Test 46
Question 450
At Bridgewater Logistics, a transformation sponsor lists ten AI ideas but has funding for one. Evidence: Ideas differ in readiness, benefits and regulatory exposure. The business goal is to prioritize with weighted criteria before funding a pilot. Only non-sensitive synthetic data and approved sandbox systems may be used. Which bounded trial with synthetic records best tests the proposed approach?
- Relabel a synthetic edge-case batch with independent reviewers
- Run paraphrased versions of the same fictional request repeatedly
- Apply a fictional scoring rubric to three competing proposals
- Test a fictional letter containing an unapproved delivery guarantee
Explanation
Correct Answer
C:
Ideas differ in readiness, benefits and regulatory exposure. Apply a fictional scoring rubric to three competing proposals. A weighted assessment exposes tradeoffs among business benefit, readiness, cost and compliance exposure. It helps choose a feasible bounded pilot rather than committing to the loudest proposal.
Incorrect Answers
A:
“Relabel a synthetic edge-case batch with independent reviewers” tests a proposed change for the condition: Training labels are inconsistent and rare classes have few examples. This question instead requires an action that can prioritize with weighted criteria before funding a pilot. Ideas differ in readiness, benefits and regulatory exposure. Apply a fictional scoring rubric to three competing proposals. A weighted assessment exposes tradeoffs among business benefit, readiness, cost and compliance exposure. It helps choose a feasible bounded pilot rather than committing to the loudest proposal.
B:
“Run paraphrased versions of the same fictional request repeatedly” tests a proposed change for the condition: Small phrasing changes cause inconsistent answers in a business-critical step. This question instead requires an action that can prioritize with weighted criteria before funding a pilot. Ideas differ in readiness, benefits and regulatory exposure. Apply a fictional scoring rubric to three competing proposals. A weighted assessment exposes tradeoffs among business benefit, readiness, cost and compliance exposure. It helps choose a feasible bounded pilot rather than committing to the loudest proposal.
D:
“Test a fictional letter containing an unapproved delivery guarantee” tests a proposed change for the condition: The proposed wording could create commitments that the legal owner has not approved. This question instead requires an action that can prioritize with weighted criteria before funding a pilot. Ideas differ in readiness, benefits and regulatory exposure. Apply a fictional scoring rubric to three competing proposals. A weighted assessment exposes tradeoffs among business benefit, readiness, cost and compliance exposure. It helps choose a feasible bounded pilot rather than committing to the loudest proposal.
Question 451
At Redwood Consulting, an agent is halfway through preparing a multi-step report when a reviewer objects. Evidence: Restarting could duplicate work and bypass reviewed steps. The business goal is to use appropriate task state and human feedback rather than blind restart. The sponsor needs an evidence-based diagnosis rather than a guessed fix. Which initial investigation best identifies the controlling constraint before any solution is chosen?
- Inspect existing task progress, artifacts and pending approval
- Inspect where tokens are stored and which callers can reach the endpoint
- List connector privileges, approval owners and audit requirements
- Confirm app, identity, license and device management policy
Explanation
Correct Answer
A:
Restarting could duplicate work and bypass reviewed steps. Inspect existing task progress, artifacts and pending approval.
Incorrect Answers
B:
“Inspect where tokens are stored and which callers can reach the endpoint” is a diagnostic step for the reported condition: Exposed credentials could be reused for unauthorized inference. This question instead requires an action that can use appropriate task state and human feedback rather than blind restart. Restarting could duplicate work and bypass reviewed steps. Inspect existing task progress, artifacts and pending approval.
C:
“List connector privileges, approval owners and audit requirements” is a diagnostic step for the reported condition: Reading policy is different from invoking a high-impact connector action. This question instead requires an action that can use appropriate task state and human feedback rather than blind restart. Restarting could duplicate work and bypass reviewed steps. Inspect existing task progress, artifacts and pending approval.
D:
“Confirm app, identity, license and device management policy” is a diagnostic step for the reported condition: The desktop-only instructions may not describe mobile access or capabilities. This question instead requires an action that can use appropriate task state and human feedback rather than blind restart. Restarting could duplicate work and bypass reviewed steps. Inspect existing task progress, artifacts and pending approval.
Question 452
At Sapphire Mobility, a manufacturer wants to predict rare equipment failures from sensor histories. Evidence: Threshold rules miss complex multi-variable patterns. The business goal is to use predictive ML when labeled trends add real signal. The sponsor needs an evidence-based diagnosis rather than a guessed fix. Which initial investigation best identifies the controlling constraint before any solution is chosen?
- Inspect language distribution of indexes and failed queries
- Verify sensors, failure labels and class imbalance
- Review ranking failures with labeled relevance judgments
- Calculate impact per error and proposed action volume
Explanation
Correct Answer
B:
Threshold rules miss complex multi-variable patterns. Verify sensors, failure labels and class imbalance. When multiple signals interact, a fixed threshold may lose useful predictive information. Evaluate a labeled model against the current rule baseline using task-relevant false-positive and false-negative costs.
Incorrect Answers
A:
“Inspect language distribution of indexes and failed queries” is a diagnostic step for the reported condition: Vector similarity may degrade for underrepresented languages. This question instead requires an action that can use predictive ML when labeled trends add real signal. Threshold rules miss complex multi-variable patterns. Verify sensors, failure labels and class imbalance. When multiple signals interact, a fixed threshold may lose useful predictive information. Evaluate a labeled model against the current rule baseline using task-relevant false-positive and false-negative costs.
C:
“Review ranking failures with labeled relevance judgments” is a diagnostic step for the reported condition: Initial keyword matching is broad and top result quality is poor. This question instead requires an action that can use predictive ML when labeled trends add real signal. Threshold rules miss complex multi-variable patterns. Verify sensors, failure labels and class imbalance. When multiple signals interact, a fixed threshold may lose useful predictive information. Evaluate a labeled model against the current rule baseline using task-relevant false-positive and false-negative costs.
D:
“Calculate impact per error and proposed action volume” is a diagnostic step for the reported condition: One incorrect dispatch date could be repeated to thousands of customers. This question instead requires an action that can use predictive ML when labeled trends add real signal. Threshold rules miss complex multi-variable patterns. Verify sensors, failure labels and class imbalance. When multiple signals interact, a fixed threshold may lose useful predictive information. Evaluate a labeled model against the current rule baseline using task-relevant false-positive and false-negative costs.
Question 453
At Clearwater Partners, an organization imposes the same lengthy review on all AI use cases. Evidence: A low-risk note formatter and high-stakes hiring scorer carry different risks. The business goal is to scale governance depth to documented impact and autonomy. The proposed solution must respect documented authority and organizational controls. Which approach most directly meets the stated business objective within the approved authority?
- Name a sponsor and communicate approved use cases and expectations
- Evaluate and mitigate proxy-driven disparities with approved methods
- Use only approved datasets with documented rights and minimization
- Define tiered release requirements with non-negotiable safety baselines
Explanation
Correct Answer
D:
A low-risk note formatter and high-stakes hiring scorer carry different risks. Define tiered release requirements with non-negotiable safety baselines.
Incorrect Answers
A:
“Name a sponsor and communicate approved use cases and expectations” is a design or implementation response to the condition: Employees receive mixed signals about priority and approved use. This question instead requires an action that can scale governance depth to documented impact and autonomy. A low-risk note formatter and high-stakes hiring scorer carry different risks. Define tiered release requirements with non-negotiable safety baselines.
B:
“Evaluate and mitigate proxy-driven disparities with approved methods” is a design or implementation response to the condition: Proxies may create discriminatory patterns even without explicit protected fields. This question instead requires an action that can scale governance depth to documented impact and autonomy. A low-risk note formatter and high-stakes hiring scorer carry different risks. Define tiered release requirements with non-negotiable safety baselines.
C:
“Use only approved datasets with documented rights and minimization” is a design or implementation response to the condition: Rights to read documents do not automatically permit model training. This question instead requires an action that can scale governance depth to documented impact and autonomy. A low-risk note formatter and high-stakes hiring scorer carry different risks. Define tiered release requirements with non-negotiable safety baselines.
Question 454
At Solstice Retail, a multinational enterprise launches AI pilots in several departments without common oversight. Evidence: Different owners approve conflicting risk exceptions. The business goal is to create cross-functional governance accountability. The operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor champion misinformation and coaching consistency
- Review ungoverned pilots and overdue policy exceptions
- Monitor overdue remediation and misleading compliance metrics
- Monitor missing logs and retention policy breaches
Explanation
Correct Answer
B:
Different owners approve conflicting risk exceptions. Review ungoverned pilots and overdue policy exceptions. Named cross-functional decision rights and escalation paths prevent departments from independently granting incompatible AI-risk exceptions.
Incorrect Answers
A:
“Monitor champion misinformation and coaching consistency” monitors this signal: Some champions share risky shortcuts and outdated feature claims. This question instead requires an action that can create cross-functional governance accountability. Different owners approve conflicting risk exceptions. Review ungoverned pilots and overdue policy exceptions. Named cross-functional decision rights and escalation paths prevent departments from independently granting incompatible AI-risk exceptions.
C:
“Monitor overdue remediation and misleading compliance metrics” monitors this signal: No evidence links governance to fewer incidents or better outcomes. This question instead requires an action that can create cross-functional governance accountability. Different owners approve conflicting risk exceptions. Review ungoverned pilots and overdue policy exceptions. Named cross-functional decision rights and escalation paths prevent departments from independently granting incompatible AI-risk exceptions.
D:
“Monitor missing logs and retention policy breaches” monitors this signal: Reviewers cannot reconstruct who authorized the change. This question instead requires an action that can create cross-functional governance accountability. Different owners approve conflicting risk exceptions. Review ungoverned pilots and overdue policy exceptions. Named cross-functional decision rights and escalation paths prevent departments from independently granting incompatible AI-risk exceptions.
Question 455
At Pinecrest Services, a school proposes uploading identifiable student photos to test a vision model. Evidence: A less sensitive evaluation could answer the technical question. The business goal is to minimize identifiable image exposure in model pilots. The operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor image retention and unauthorized sample reuse
- Monitor stale work context and access-control changes
- Monitor citation age and style regression independently
- Monitor hallucinated visual details and multimodal charges
Explanation
Correct Answer
A:
A less sensitive evaluation could answer the technical question. Monitor image retention and unauthorized sample reuse. Using actual personal images is unnecessary when a synthetic or redacted sample can test the visual capability. Reducing source sensitivity limits the blast radius of a pilot failure.
Incorrect Answers
B:
“Monitor stale work context and access-control changes” monitors this signal: The answer depends on authorized Microsoft 365 work context. This question instead requires an action that can minimize identifiable image exposure in model pilots. A less sensitive evaluation could answer the technical question. Monitor image retention and unauthorized sample reuse. Using actual personal images is unnecessary when a synthetic or redacted sample can test the visual capability. Reducing source sensitivity limits the blast radius of a pilot failure.
C:
“Monitor citation age and style regression independently” monitors this signal: Fresh facts and style behavior are separate concerns. This question instead requires an action that can minimize identifiable image exposure in model pilots. A less sensitive evaluation could answer the technical question. Monitor image retention and unauthorized sample reuse. Using actual personal images is unnecessary when a synthetic or redacted sample can test the visual capability. Reducing source sensitivity limits the blast radius of a pilot failure.
D:
“Monitor hallucinated visual details and multimodal charges” monitors this signal: Text-only models cannot inspect image evidence. This question instead requires an action that can minimize identifiable image exposure in model pilots. A less sensitive evaluation could answer the technical question. Monitor image retention and unauthorized sample reuse. Using actual personal images is unnecessary when a synthetic or redacted sample can test the visual capability. Reducing source sensitivity limits the blast radius of a pilot failure.
Question 456
At Juniper Operations, a project team evaluates an AI agent only for answer accuracy. Evidence: The agent can call tools and access internal data repositories. The business goal is to threat-model AI application and tool boundaries. The proposed solution must respect documented authority and organizational controls. Which approach most directly meets the stated business objective within the approved authority?
- Generate reviewable reason categories and appeal instructions
- Set budgets and review usage through supported cost management
- Design least privilege, isolation, authentication and audit controls
- Prevent unsafe advice and route uncertain cases to trained staff
Explanation
Correct Answer
C:
The agent can call tools and access internal data repositories. Design least privilege, isolation, authentication and audit controls.
Incorrect Answers
A:
“Generate reviewable reason categories and appeal instructions” is a design or implementation response to the condition: Customers need a comprehensible explanation and correction mechanism. This question instead requires an action that can threat-model AI application and tool boundaries. The agent can call tools and access internal data repositories. Design least privilege, isolation, authentication and audit controls.
B:
“Set budgets and review usage through supported cost management” is a design or implementation response to the condition: Consumption can rise faster than the annual forecast. This question instead requires an action that can threat-model AI application and tool boundaries. The agent can call tools and access internal data repositories. Design least privilege, isolation, authentication and audit controls.
D:
“Prevent unsafe advice and route uncertain cases to trained staff” is a design or implementation response to the condition: Incorrect advice could cause physical harm. This question instead requires an action that can threat-model AI application and tool boundaries. The agent can call tools and access internal data repositories. Design least privilege, isolation, authentication and audit controls.
Question 457
At Northlake Services, an international service desk wants one summarizer for multilingual requests. Evidence: Evaluation data covers only English and may hide translation losses. The business goal is to validate language capability on actual supported locales. The proposed solution must respect documented authority and organizational controls. Which approach most directly meets the stated business objective within the approved authority?
- Publish a documented ranking and choose a feasible limited first trial
- Use original brand-approved prompts and human creative review
- Compare eligible models using the same approved benchmarks and terms
- Pilot with multilingual human-reviewed examples and escalation for unsupported languages
Explanation
Correct Answer
D:
Evaluation data covers only English and may hide translation losses. Pilot with multilingual human-reviewed examples and escalation for unsupported languages.
Incorrect Answers
A:
“Publish a documented ranking and choose a feasible limited first trial” tests a proposed change for the condition: Ideas differ in readiness, benefits and regulatory exposure. This question instead requires an action that can validate language capability on actual supported locales. Evaluation data covers only English and may hide translation losses. Pilot with multilingual human-reviewed examples and escalation for unsupported languages.
B:
“Use original brand-approved prompts and human creative review” is a design or implementation response to the condition: Similar branded content may create legal or reputational risks. This question instead requires an action that can validate language capability on actual supported locales. Evaluation data covers only English and may hide translation losses. Pilot with multilingual human-reviewed examples and escalation for unsupported languages.
C:
“Compare eligible models using the same approved benchmarks and terms” is a design or implementation response to the condition: Contractual hosting, customization and governance requirements differ. This question instead requires an action that can validate language capability on actual supported locales. Evaluation data covers only English and may hide translation losses. Pilot with multilingual human-reviewed examples and escalation for unsupported languages.
Question 458
At Elmwood Banking, a company announces AI tools once and assumes adoption follows. Evidence: Teams miss updates and do not understand why workflows change. The business goal is to plan repeated role-targeted change communications. The operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor recurring open links and unsupported license states
- Monitor unresolved incidents and response time to feedback
- Monitor misclassified high-risk uses and policy bypass
- Monitor knowledge retention and inconsistent manager guidance
Explanation
Correct Answer
D:
Teams miss updates and do not understand why workflows change. Monitor knowledge retention and inconsistent manager guidance. Workers need to know which tasks change, which controls remain and where to report failures. Repeated role-specific demonstrations and feedback channels reduce adoption confusion better than a one-off launch email.
Incorrect Answers
A:
“Monitor recurring open links and unsupported license states” monitors this signal: Enterprise repositories contain broad anonymous sharing links. This question instead requires an action that can plan repeated role-targeted change communications. Teams miss updates and do not understand why workflows change. Monitor knowledge retention and inconsistent manager guidance. Workers need to know which tasks change, which controls remain and where to report failures. Repeated role-specific demonstrations and feedback channels reduce adoption confusion better than a one-off launch email.
B:
“Monitor unresolved incidents and response time to feedback” monitors this signal: Issues are visible only to individual users and are not triaged. This question instead requires an action that can plan repeated role-targeted change communications. Teams miss updates and do not understand why workflows change. Monitor knowledge retention and inconsistent manager guidance. Workers need to know which tasks change, which controls remain and where to report failures. Repeated role-specific demonstrations and feedback channels reduce adoption confusion better than a one-off launch email.
C:
“Monitor misclassified high-risk uses and policy bypass” monitors this signal: A low-risk note formatter and high-stakes hiring scorer carry different risks. This question instead requires an action that can plan repeated role-targeted change communications. Teams miss updates and do not understand why workflows change. Monitor knowledge retention and inconsistent manager guidance. Workers need to know which tasks change, which controls remain and where to report failures. Repeated role-specific demonstrations and feedback channels reduce adoption confusion better than a one-off launch email.
Question 459
At Coral Bay Systems, a CFO estimates AI expense from model tokens alone. Evidence: Integration, human review, logging and employee training are omitted. The business goal is to include operational costs beyond the advertised model rate. The operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor actual spend versus TCO categories and revise forecasts
- Monitor vendor policy changes and data upload exceptions
- Monitor realized savings, rework and hidden supervision expense
- Monitor retrieval precision and drifting embedding models
Explanation
Correct Answer
A:
Integration, human review, logging and employee training are omitted. Monitor actual spend versus TCO categories and revise forecasts.
Incorrect Answers
B:
“Monitor vendor policy changes and data upload exceptions” monitors this signal: The vendor retention and training terms were not evaluated. This question instead requires an action that can include operational costs beyond the advertised model rate. Integration, human review, logging and employee training are omitted. Monitor actual spend versus TCO categories and revise forecasts.
C:
“Monitor realized savings, rework and hidden supervision expense” monitors this signal: No baseline handling time, error rate or deployment cost is recorded. This question instead requires an action that can include operational costs beyond the advertised model rate. Integration, human review, logging and employee training are omitted. Monitor actual spend versus TCO categories and revise forecasts.
D:
“Monitor retrieval precision and drifting embedding models” monitors this signal: Similarity matching is distinct from composing a final recommendation. This question instead requires an action that can include operational costs beyond the advertised model rate. Integration, human review, logging and employee training are omitted. Monitor actual spend versus TCO categories and revise forecasts.
Question 460
At Solstice Retail, a regional support search finds English documents but misses equivalent Spanish guidance. Evidence: Vector similarity may degrade for underrepresented languages. The business goal is to test cross-language retrieval empirically. The operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor opaque rankings and unexplained score swings
- Monitor OCR failures and source format changes
- Monitor per-language retrieval quality and translation drift
- Monitor false visual assertions and camera-domain drift
Explanation
Correct Answer
C:
Vector similarity may degrade for underrepresented languages. Monitor per-language retrieval quality and translation drift.
Incorrect Answers
A:
“Monitor opaque rankings and unexplained score swings” monitors this signal: The decision-maker must justify awards to an audit committee. This question instead requires an action that can test cross-language retrieval empirically. Vector similarity may degrade for underrepresented languages. Monitor per-language retrieval quality and translation drift.
B:
“Monitor OCR failures and source format changes” monitors this signal: Corrupt source extraction can make even a grounded answer wrong. This question instead requires an action that can test cross-language retrieval empirically. Vector similarity may degrade for underrepresented languages. Monitor per-language retrieval quality and translation drift.
D:
“Monitor false visual assertions and camera-domain drift” monitors this signal: Text descriptions alone omit visual indicators needed for triage. This question instead requires an action that can test cross-language retrieval empirically. Vector similarity may degrade for underrepresented languages. Monitor per-language retrieval quality and translation drift.