Microsoft AB-731 Practice Test 94
Question 930
At Evergreen Research, a finance assistant answers a query and is also asked to transfer funds. Evidence: Retrieving an account statement does not authorize transactions. The business goal is to separate read-only grounding from high-impact actions. 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 service terms, sensitivity classification and intended storage
- Review endpoint dependencies, timeouts and critical workflow priority
- Trace action permissions, user identity and approval workflow
- Compare the transcript to source audio and identify unclear speakers
Explanation
Correct Answer
C:
Retrieving an account statement does not authorize transactions. Trace action permissions, user identity and approval workflow.
Incorrect Answers
A:
“Inspect service terms, sensitivity classification and intended storage” is a diagnostic step for the reported condition: The vendor retention and training terms were not evaluated. This question instead requires an action that can separate read-only grounding from high-impact actions. Retrieving an account statement does not authorize transactions. Trace action permissions, user identity and approval workflow.
B:
“Review endpoint dependencies, timeouts and critical workflow priority” is a diagnostic step for the reported condition: A single dependent service outage stops all customer help. This question instead requires an action that can separate read-only grounding from high-impact actions. Retrieving an account statement does not authorize transactions. Trace action permissions, user identity and approval workflow.
D:
“Compare the transcript to source audio and identify unclear speakers” is a diagnostic step for the reported condition: Poor transcription can propagate incorrect speaker attribution into summaries. This question instead requires an action that can separate read-only grounding from high-impact actions. Retrieving an account statement does not authorize transactions. Trace action permissions, user identity and approval workflow.
Question 931
At Crownridge Travel, a CFO wants to provide premium Copilot seats to every employee automatically. Evidence: Usage patterns vary and some workflows may use included chat capabilities. The business goal is to compare entitlements and cost against actual user needs. 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?
- Identify how users report issues and how severe they are
- Identify baseline effort, incremental expenses and impact metrics
- Inventory licenses, role requirements and expected daily tasks
- Identify which evidence and criteria support the recommendation
Explanation
Correct Answer
C:
Usage patterns vary and some workflows may use included chat capabilities. Inventory licenses, role requirements and expected daily tasks.
Incorrect Answers
A:
“Identify how users report issues and how severe they are” is a diagnostic step for the reported condition: Issues are visible only to individual users and are not triaged. This question instead requires an action that can compare entitlements and cost against actual user needs. Usage patterns vary and some workflows may use included chat capabilities. Inventory licenses, role requirements and expected daily tasks.
B:
“Identify baseline effort, incremental expenses and impact metrics” is a diagnostic step for the reported condition: Real costs include licenses, training, evaluation and human review. This question instead requires an action that can compare entitlements and cost against actual user needs. Usage patterns vary and some workflows may use included chat capabilities. Inventory licenses, role requirements and expected daily tasks.
D:
“Identify which evidence and criteria support the recommendation” is a diagnostic step for the reported condition: Customers need a comprehensible explanation and correction mechanism. This question instead requires an action that can compare entitlements and cost against actual user needs. Usage patterns vary and some workflows may use included chat capabilities. Inventory licenses, role requirements and expected daily tasks.
Question 932
At Blue Ridge Technology, a utility customer asks a bot to negotiate a nonstandard billing adjustment. Evidence: The request is beyond published policy and automated authority. The business goal is to provide human escalation from governed agents. The proposed solution must respect documented authority and organizational controls. Which approach most directly meets the stated business objective within the approved authority?
- Test a suitable existing model before considering custom training
- Deploy a prebuilt option if it meets scope and licensing requirements
- Define workflow instructions independent of unsupported model toggles
- Route exception requests with sufficient context to a trained employee
Explanation
Correct Answer
D:
The request is beyond published policy and automated authority. Route exception requests with sufficient context to a trained employee.
Incorrect Answers
A:
“Test a suitable existing model before considering custom training” is a design or implementation response to the condition: The available budget and data are far smaller than required for foundation training. This question instead requires an action that can provide human escalation from governed agents. The request is beyond published policy and automated authority. Route exception requests with sufficient context to a trained employee.
B:
“Deploy a prebuilt option if it meets scope and licensing requirements” is a design or implementation response to the condition: A prebuilt agent may already support search without bespoke code. This question instead requires an action that can provide human escalation from governed agents. The request is beyond published policy and automated authority. Route exception requests with sufficient context to a trained employee.
C:
“Define workflow instructions independent of unsupported model toggles” is a design or implementation response to the condition: The supported surface and administrative policies govern model availability. This question instead requires an action that can provide human escalation from governed agents. The request is beyond published policy and automated authority. Route exception requests with sufficient context to a trained employee.
Question 933
At Pinecrest Services, 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 operating team needs an ongoing check after the initial rollout. Which post-rollout signal is most important for detecting regression or governance drift?
- Monitor query scope expansion and inappropriate personnel inference
- Monitor false assertions about connected data access
- Monitor repeated tasks and lost approval checkpoints
- Monitor changed access lists and index filter bypass
Explanation
Correct Answer
C:
Restarting could duplicate work and bypass reviewed steps. Monitor repeated tasks and lost approval checkpoints.
Incorrect Answers
A:
“Monitor query scope expansion and inappropriate personnel inference” monitors this signal: Graph relationships may expose sensitive work patterns and restricted files. 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. Monitor repeated tasks and lost approval checkpoints.
B:
“Monitor false assertions about connected data access” monitors this signal: The service has not been authorized to access the requested private file. 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. Monitor repeated tasks and lost approval checkpoints.
D:
“Monitor changed access lists and index filter bypass” monitors this signal: Index query results lack caller-specific authorization filters. 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. Monitor repeated tasks and lost approval checkpoints.
Question 934
At Meridian Imports, an IT leader wants to renew hundreds of Copilot seats based only on logins. Evidence: License usage is not equivalent to measurable task improvement. The business goal is to evaluate per-seat productivity and quality before procurement. The project owner requires a measurable acceptance criterion before signoff. Which evidence would most convincingly demonstrate that the requirement has been met?
- Unapproved agents remain unavailable to production users
- Spend decisions reflect measured benefit rather than raw adoption
- Routine answers cite current policy and exceptions escalate
- Observed unit cost matches forecast within tolerance
Explanation
Correct Answer
B:
License usage is not equivalent to measurable task improvement. Spend decisions reflect measured benefit rather than raw adoption.
Incorrect Answers
A:
“Unapproved agents remain unavailable to production users” validates an outcome in a case where the evidence is: Unreviewed deployment could expose confidential data or unauthorized tools. This question instead requires an action that can evaluate per-seat productivity and quality before procurement. License usage is not equivalent to measurable task improvement. Spend decisions reflect measured benefit rather than raw adoption.
C:
“Routine answers cite current policy and exceptions escalate” validates an outcome in a case where the evidence is: The questions recur and answers must reflect maintained policy knowledge. This question instead requires an action that can evaluate per-seat productivity and quality before procurement. License usage is not equivalent to measurable task improvement. Spend decisions reflect measured benefit rather than raw adoption.
D:
“Observed unit cost matches forecast within tolerance” validates an outcome in a case where the evidence is: Input and output tokens and deployment type all affect billing. This question instead requires an action that can evaluate per-seat productivity and quality before procurement. License usage is not equivalent to measurable task improvement. Spend decisions reflect measured benefit rather than raw adoption.
Question 935
At Coral Bay Systems, a public chatbot prototype occasionally outputs unsafe or abusive language. Evidence: The release needs safety checks beyond normal accuracy tests. The business goal is to integrate suitable content safety screening into deployment criteria. 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?
- Estimate dataset scale, compute cost and target performance
- Identify relevant harm categories and platform filtering limits
- Inspect transcript availability and speaker/decision confirmation
- Record token volumes, retries and actual chosen billing meters
Explanation
Correct Answer
B:
The release needs safety checks beyond normal accuracy tests. Identify relevant harm categories and platform filtering limits. Conventional task accuracy does not measure unsafe content, tool misuse or privacy leakage. Add appropriate adversarial and policy tests as a separate release gate.
Incorrect Answers
A:
“Estimate dataset scale, compute cost and target performance” is a diagnostic step for the reported condition: The available budget and data are far smaller than required for foundation training. This question instead requires an action that can integrate suitable content safety screening into deployment criteria. The release needs safety checks beyond normal accuracy tests. Identify relevant harm categories and platform filtering limits. Conventional task accuracy does not measure unsafe content, tool misuse or privacy leakage. Add appropriate adversarial and policy tests as a separate release gate.
C:
“Inspect transcript availability and speaker/decision confirmation” is a diagnostic step for the reported condition: Transcript includes suggestions that were not formally accepted. This question instead requires an action that can integrate suitable content safety screening into deployment criteria. The release needs safety checks beyond normal accuracy tests. Identify relevant harm categories and platform filtering limits. Conventional task accuracy does not measure unsafe content, tool misuse or privacy leakage. Add appropriate adversarial and policy tests as a separate release gate.
D:
“Record token volumes, retries and actual chosen billing meters” is a diagnostic step for the reported condition: Input and output tokens and deployment type all affect billing. This question instead requires an action that can integrate suitable content safety screening into deployment criteria. The release needs safety checks beyond normal accuracy tests. Identify relevant harm categories and platform filtering limits. Conventional task accuracy does not measure unsafe content, tool misuse or privacy leakage. Add appropriate adversarial and policy tests as a separate release gate.
Question 936
At Silverpine Group, a payroll unit proposes rewriting numerical reconciliation checks as chatbot reasoning. Evidence: The reconciliation is deterministic arithmetic with known control rules. The business goal is to avoid generative substitution for reliable rule-based computation. 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 AI suggestions that conflict with accounting controls
- Monitor missing traces, excess retention and alert thresholds
- Monitor scope creep and benefit erosion during scale-up
- Monitor transcription errors, consent settings and missing audio segments
Explanation
Correct Answer
A:
The reconciliation is deterministic arithmetic with known control rules. Monitor AI suggestions that conflict with accounting controls. Known reconciliation rules are deterministic, so they should be implemented as repeatable arithmetic with explicit exception handling. A generative response can explain results but should not substitute for controlled calculation.
Incorrect Answers
B:
“Monitor missing traces, excess retention and alert thresholds” monitors this signal: There is no trace from answer to model version and retrieved sources. This question instead requires an action that can avoid generative substitution for reliable rule-based computation. The reconciliation is deterministic arithmetic with known control rules. Monitor AI suggestions that conflict with accounting controls. Known reconciliation rules are deterministic, so they should be implemented as repeatable arithmetic with explicit exception handling. A generative response can explain results but should not substitute for controlled calculation.
C:
“Monitor scope creep and benefit erosion during scale-up” monitors this signal: High deployment effort conceals whether one use case can succeed. This question instead requires an action that can avoid generative substitution for reliable rule-based computation. The reconciliation is deterministic arithmetic with known control rules. Monitor AI suggestions that conflict with accounting controls. Known reconciliation rules are deterministic, so they should be implemented as repeatable arithmetic with explicit exception handling. A generative response can explain results but should not substitute for controlled calculation.
D:
“Monitor transcription errors, consent settings and missing audio segments” monitors this signal: Poor transcription can propagate incorrect speaker attribution into summaries. This question instead requires an action that can avoid generative substitution for reliable rule-based computation. The reconciliation is deterministic arithmetic with known control rules. Monitor AI suggestions that conflict with accounting controls. Known reconciliation rules are deterministic, so they should be implemented as repeatable arithmetic with explicit exception handling. A generative response can explain results but should not substitute for controlled calculation.
Question 937
At BrightPath Media, a logistics team proposes auto-sending AI-generated shipping notices. Evidence: One incorrect dispatch date could be repeated to thousands of customers. The business goal is to consider error amplification when estimating AI business value. 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?
- Review endpoint dependencies, timeouts and critical workflow priority
- Calculate impact per error and proposed action volume
- Identify hosting, retrieval, licensing, change and supervision cost buckets
- Compare responder and non-responder populations and resolution rates
Explanation
Correct Answer
B:
One incorrect dispatch date could be repeated to thousands of customers. Calculate impact per error and proposed action volume.
Incorrect Answers
A:
“Review endpoint dependencies, timeouts and critical workflow priority” is a diagnostic step for the reported condition: A single dependent service outage stops all customer help. This question instead requires an action that can consider error amplification when estimating AI business value. One incorrect dispatch date could be repeated to thousands of customers. Calculate impact per error and proposed action volume.
C:
“Identify hosting, retrieval, licensing, change and supervision cost buckets” is a diagnostic step for the reported condition: Integration, human review, logging and employee training are omitted. This question instead requires an action that can consider error amplification when estimating AI business value. One incorrect dispatch date could be repeated to thousands of customers. Calculate impact per error and proposed action volume.
D:
“Compare responder and non-responder populations and resolution rates” is a diagnostic step for the reported condition: Feedback is self-selected and omits escalated failures. This question instead requires an action that can consider error amplification when estimating AI business value. One incorrect dispatch date could be repeated to thousands of customers. Calculate impact per error and proposed action volume.
Question 938
At Redwood Consulting, an executive declares an AI pilot successful because weekly users doubled. Evidence: No customer quality, time savings or cost data were collected. The business goal is to define success through business outcomes rather than activity. The proposed solution must respect documented authority and organizational controls. Which approach most directly meets the stated business objective within the approved authority?
- Set outcome targets and collect matched pilot comparisons
- Remediate high-risk data access and confirm pilot prerequisites
- Define stop/rollback, notification and correction procedures
- Build transparent ROI reporting with benefit attribution
Explanation
Correct Answer
A:
No customer quality, time savings or cost data were collected. Set outcome targets and collect matched pilot comparisons.
Incorrect Answers
B:
“Remediate high-risk data access and confirm pilot prerequisites” is a design or implementation response to the condition: Enterprise repositories contain broad anonymous sharing links. This question instead requires an action that can define success through business outcomes rather than activity. No customer quality, time savings or cost data were collected. Set outcome targets and collect matched pilot comparisons.
C:
“Define stop/rollback, notification and correction procedures” is a design or implementation response to the condition: There is no owner or process for AI-specific incidents. This question instead requires an action that can define success through business outcomes rather than activity. No customer quality, time savings or cost data were collected. Set outcome targets and collect matched pilot comparisons.
D:
“Build transparent ROI reporting with benefit attribution” is a design or implementation response to the condition: Real costs include licenses, training, evaluation and human review. This question instead requires an action that can define success through business outcomes rather than activity. No customer quality, time savings or cost data were collected. Set outcome targets and collect matched pilot comparisons.
Question 939
At Northlake Services, a leadership team schedules AI deployment before verifying source permissions. Evidence: Enterprise repositories contain broad anonymous sharing links. The business goal is to run readiness checks before connecting work-grounded assistants. 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 workload and uneven team coverage
- Monitor unresolved complaints and disparate correction times
- Monitor recurring open links and unsupported license states
- Monitor support load and quality loss with expanded usage
Explanation
Correct Answer
C:
Enterprise repositories contain broad anonymous sharing links. Monitor recurring open links and unsupported license states.
Incorrect Answers
A:
“Monitor champion workload and uneven team coverage” monitors this signal: Local help and peer examples are not available. This question instead requires an action that can run readiness checks before connecting work-grounded assistants. Enterprise repositories contain broad anonymous sharing links. Monitor recurring open links and unsupported license states.
B:
“Monitor unresolved complaints and disparate correction times” monitors this signal: Users may be unable to correct harmful mistakes. This question instead requires an action that can run readiness checks before connecting work-grounded assistants. Enterprise repositories contain broad anonymous sharing links. Monitor recurring open links and unsupported license states.
D:
“Monitor support load and quality loss with expanded usage” monitors this signal: Support capacity and user task quality have not been proven. This question instead requires an action that can run readiness checks before connecting work-grounded assistants. Enterprise repositories contain broad anonymous sharing links. Monitor recurring open links and unsupported license states.
Question 940
At Tamarack Logistics, an AI pilot claims 100 percent success after five easy tests. Evidence: The sample cannot represent real volume and edge cases. The business goal is to design statistically meaningful evaluations for the use case. The project owner requires a measurable acceptance criterion before signoff. Which evidence would most convincingly demonstrate that the requirement has been met?
- Answers switch to the new approved clause with version citation
- Output quotes only allowed material and names missing context
- The investment case shows net value under realistic cost assumptions
- Results report uncertainty and fail safely when data is insufficient
Explanation
Correct Answer
D:
The sample cannot represent real volume and edge cases. Results report uncertainty and fail safely when data is insufficient. A small convenience sample cannot reveal tail failures or rare operational cases. Plan representative coverage, required confidence, relevant slices and escalation before using results for rollout decisions.
Incorrect Answers
A:
“Answers switch to the new approved clause with version citation” validates an outcome in a case where the evidence is: The knowledge index was last refreshed before the recent policy revision. This question instead requires an action that can design statistically meaningful evaluations for the use case. The sample cannot represent real volume and edge cases. Results report uncertainty and fail safely when data is insufficient. A small convenience sample cannot reveal tail failures or rare operational cases. Plan representative coverage, required confidence, relevant slices and escalation before using results for rollout decisions.
B:
“Output quotes only allowed material and names missing context” validates an outcome in a case where the evidence is: The user can access the approved summary but not full agreement text. This question instead requires an action that can design statistically meaningful evaluations for the use case. The sample cannot represent real volume and edge cases. Results report uncertainty and fail safely when data is insufficient. A small convenience sample cannot reveal tail failures or rare operational cases. Plan representative coverage, required confidence, relevant slices and escalation before using results for rollout decisions.
C:
“The investment case shows net value under realistic cost assumptions” validates an outcome in a case where the evidence is: Integration, human review, logging and employee training are omitted. This question instead requires an action that can design statistically meaningful evaluations for the use case. The sample cannot represent real volume and edge cases. Results report uncertainty and fail safely when data is insufficient. A small convenience sample cannot reveal tail failures or rare operational cases. Plan representative coverage, required confidence, relevant slices and escalation before using results for rollout decisions.