Administering Information Security in Microsoft 365 SC-401 Practice Test 79
Question 780
A policy is required to identify elevated insider risk when employees browse AI sites and submit confidential prompts. After configuring the selected control, what acceptance evidence MOST DIRECTLY establishes that the stated boundary works?
- Test supported sensitive prompts create the intended risk events for a scoped user
- An AI data security dashboard is enabled but has no validated workload or interaction evidence
- A document is labeled Confidential but access by an excluded user was not assessed
- Copilot licenses are assigned, but risky content access and audit events have not been evaluated
Explanation
Correct Answer
A:Test supported sensitive prompts create the intended risk events for a scoped user. This observation verifies the required control: Configure Insider Risk Management risky AI usage detection with its required supported indicators and extensions. Risky AI usage policies correlate eligible browser or AI interaction signals with user risk.
Incorrect Answers
B:A dashboard toggle alone is not observed protection or monitoring. In the present case the decisive acceptance evidence is: Test supported sensitive prompts create the intended risk events for a scoped user
C:Classification name alone does not establish authorized AI data access. In the present case the decisive acceptance evidence is: Test supported sensitive prompts create the intended risk events for a scoped user
D:Entitlement does not attest to AI data security outcome. In the present case the decisive acceptance evidence is: Test supported sensitive prompts create the intended risk events for a scoped user
Question 781
A company has deployed custom agents and wants monitoring when an agent shares priority SharePoint files externally. Which Microsoft Purview configuration or security administration decision BEST meets the requirement?
- Create an Exchange mail rule for human messages only
- Rely solely on the agent's friendly name as a trusted identity
- Disable all SharePoint audit activity to reduce alerts
- Use the relevant Insider Risk Management risky agents detection and supported agent data-risk signals
Explanation
Correct Answer
D:Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage. This choice directly addresses the stated protected-data workflow.
Incorrect Answers
A:Human email controls do not comprehensively detect agent file sharing. Consequently it cannot satisfy the stated requirement.
B:Agent naming provides no evidence of safe data access. Consequently it cannot satisfy the stated requirement.
C:Loss of telemetry makes agent exposure harder to investigate. Consequently it cannot satisfy the stated requirement.
Question 782
A company has deployed custom agents and wants monitoring when an agent shares priority SharePoint files externally. During the pilot, an administrator instead implements this approach: Create an Exchange mail rule for human messages only. The required outcome is not achieved. Which technical diagnosis BEST explains the failure?
- Agent naming provides no evidence of safe data access
- Human email controls do not comprehensively detect agent file sharing
- Loss of telemetry makes agent exposure harder to investigate
- Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage
Explanation
Correct Answer
B:Human email controls do not comprehensively detect agent file sharing. This identifies the failure of the specific pilot implementation.
Incorrect Answers
A:Agent naming provides no evidence of safe data access. That limitation concerns a different competing approach, not this pilot change.
C:Loss of telemetry makes agent exposure harder to investigate. This is a real constraint on a separate choice rather than the stated pilot implementation.
D:Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage. This describes the sound approach but does not diagnose why the observed alternative failed.
Question 783
A company has deployed custom agents and wants monitoring when an agent shares priority SharePoint files externally. A change reviewer proposes Rely solely on the agent's friendly name as a trusted identity instead of Use the relevant Insider Risk Management risky agents detection and supported agent data-risk signals. Which assessment BEST explains the proposed design gap?
- Agent naming provides no evidence of safe data access
- Human email controls do rather than comprehensively detect agent file sharing
- Loss of telemetry makes agent exposure harder to investigate
- Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage
Explanation
Correct Answer
A:Agent naming provides no evidence of safe data access. This deficiency directly applies to the suggested competing design.
Incorrect Answers
B:Human email controls do not comprehensively detect agent file sharing. This limitation relates to a different rejected configuration.
C:Loss of telemetry makes agent exposure harder to investigate. This explains a different option rather than the actual proposal.
D:Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage. This supports the preferred design rather than evaluating the proposed alternative.
Question 784
A company has deployed custom agents and wants monitoring when an agent shares priority SharePoint files externally. After configuring the selected control, what acceptance evidence MOST DIRECTLY establishes that the stated boundary works?
- An AI data security dashboard is enabled but has no validated workload or interaction evidence
- A document is labeled Confidential but access by an excluded user was not assessed
- A test risky agent share is detected and links to the affected sensitive content
- Copilot licenses are assigned, but risky content access and audit events have not been evaluated
Explanation
Correct Answer
C:A test risky agent share is detected and links to the affected sensitive content. This observation verifies the required control: Use the relevant Insider Risk Management risky agents detection and supported agent data-risk signals. Agent-oriented risk monitoring evaluates patterns beyond ordinary user-only web usage.
Incorrect Answers
A:A dashboard toggle alone is not observed protection or monitoring. In the present case the decisive acceptance evidence is: A test risky agent share is detected and links to the affected sensitive content
B:Classification name alone does not establish authorized AI data access. In the present case the decisive acceptance evidence is: A test risky agent share is detected and links to the affected sensitive content
D:Entitlement does not attest to AI data security outcome. In the present case the decisive acceptance evidence is: A test risky agent share is detected and links to the affected sensitive content
Question 785
A project wants Microsoft Purview policies to protect confidential files used in generative AI without rewriting the existing data classification vocabulary. Which Microsoft Purview configuration or security administration decision BEST meets the requirement?
- Create a separate unsynchronized plaintext keyword list for each AI tool
- Reuse supported sensitivity labels, sensitive information types, DLP and workload permissions in the AI protection design
- Remove existing SharePoint permissions once AI is deployed
- Set every dataset to Public so Copilot can search it
Explanation
Correct Answer
B:Existing data classification and access controls can inform AI-related protection rather than duplicating definitions. This choice directly addresses the stated protected-data workflow.
Incorrect Answers
A:Disconnected lists drift from governed sensitive information and label definitions. Consequently it cannot satisfy the stated requirement.
C:Disabling access controls is the opposite of secure AI integration. Consequently it cannot satisfy the stated requirement.
D:Broad classification downgrade increases disclosure risk. Consequently it cannot satisfy the stated requirement.
Question 786
A project wants Microsoft Purview policies to protect confidential files used in generative AI without rewriting the existing data classification vocabulary. During the pilot, an administrator instead implements this approach: Create a separate unsynchronized plaintext keyword list for each AI tool. The required outcome is not achieved. Which technical diagnosis BEST explains the failure?
- Disabling access controls is the opposite of secure AI integration
- Broad classification downgrade increases disclosure risk
- Existing data classification and access controls can inform AI-related protection rather than duplicating definitions
- Disconnected lists drift from governed sensitive information and label definitions
Explanation
Correct Answer
D:Disconnected lists drift from governed sensitive information and label definitions. This identifies the failure of the specific pilot implementation.
Incorrect Answers
A:Disabling access controls is the opposite of secure AI integration. That limitation concerns a different competing approach, not this pilot change.
B:Broad classification downgrade increases disclosure risk. This is a real constraint on a separate choice rather than the stated pilot implementation.
C:Existing data classification and access controls can inform AI-related protection rather than duplicating definitions. This describes the sound approach but does not diagnose why the observed alternative failed.
Question 787
A project wants Microsoft Purview policies to protect confidential files used in generative AI without rewriting the existing data classification vocabulary. A change reviewer proposes Remove existing SharePoint permissions once AI is deployed instead of Reuse supported sensitivity labels, sensitive information types, DLP and workload permissions in the AI protection design. Which assessment BEST explains the proposed design gap?
- Disabling access controls is the opposite of secure AI integration
- Disconnected lists drift from governed sensitive information and label definitions
- Broad classification downgrade increases disclosure risk
- Existing data classification and access controls can inform AI-related protection rather than duplicating definitions
Explanation
Correct Answer
A:Disabling access controls is the opposite of secure AI integration. This deficiency directly applies to the suggested competing design.
Incorrect Answers
B:Disconnected lists drift from governed sensitive information and label definitions. This limitation relates to a different rejected configuration.
C:Broad classification downgrade increases disclosure risk. This explains a different option rather than the actual proposal.
D:Existing data classification and access controls can inform AI-related protection rather than duplicating definitions. This supports the preferred design rather than evaluating the proposed alternative.
Question 788
A project wants Microsoft Purview policies to protect confidential files used in generative AI without rewriting the existing data classification vocabulary. After configuring the selected control, what acceptance evidence MOST DIRECTLY establishes that the stated boundary works?
- An AI data security dashboard is enabled but has no validated workload or interaction evidence
- A document is labeled Confidential but access by an excluded user was not assessed
- A labeled test file retains its intended restrictions in supported AI and direct workloads
- Copilot licenses are assigned, but risky content access and audit events have not been evaluated
Explanation
Correct Answer
C:A labeled test file retains its intended restrictions in supported AI and direct workloads. This observation verifies the required control: Reuse supported sensitivity labels, sensitive information types, DLP and workload permissions in the AI protection design. Existing data classification and access controls can inform AI-related protection rather than duplicating definitions.
Incorrect Answers
A:A dashboard toggle alone is not observed protection or monitoring. In the present case the decisive acceptance evidence is: A labeled test file retains its intended restrictions in supported AI and direct workloads
B:Classification name alone does not establish authorized AI data access. In the present case the decisive acceptance evidence is: A labeled test file retains its intended restrictions in supported AI and direct workloads
D:Entitlement does not attest to AI data security outcome. In the present case the decisive acceptance evidence is: A labeled test file retains its intended restrictions in supported AI and direct workloads
Question 789
A DSPM for AI report shows a high number of sensitive files broadly accessible through SharePoint even though no AI policy violations have yet been recorded. Which Microsoft Purview configuration or security administration decision BEST meets the requirement?
- Wait for a confirmed AI leakage incident before changing broad permissions
- Suppress DSPM findings to reduce dashboard alerts
- Turn off all audit retention to reduce reporting cost
- Prioritize source sharing and permissions remediation based on exposure findings
Explanation
Correct Answer
D:Preventive posture work addresses overexposed data before a malicious AI interaction occurs. This choice directly addresses the stated protected-data workflow.
Incorrect Answers
A:A reactive-only approach leaves the documented exposure unresolved. Consequently it cannot satisfy the stated requirement.
B:Hiding risk evidence is not remediation. Consequently it cannot satisfy the stated requirement.
C:Audit lifecycle changes cannot fix SharePoint file permissions. Consequently it cannot satisfy the stated requirement.
Question 790
A DSPM for AI report shows a high number of sensitive files broadly accessible through SharePoint even though no AI policy violations have yet been recorded. During the pilot, an administrator instead implements this approach: Wait for a confirmed AI leakage incident before changing broad permissions. The required outcome is not achieved. Which technical diagnosis BEST explains the failure?
- Hiding risk evidence is not remediation
- A reactive-only approach leaves the documented exposure unresolved
- Audit lifecycle changes cannot fix SharePoint file permissions
- Preventive posture work addresses overexposed data before a malicious AI interaction occurs
Explanation
Correct Answer
B:A reactive-only approach leaves the documented exposure unresolved. This identifies the failure of the specific pilot implementation.
Incorrect Answers
A:Hiding risk evidence is not remediation. That limitation concerns a different competing approach, not this pilot change.
C:Audit lifecycle changes cannot fix SharePoint file permissions. This is a real constraint on a separate choice rather than the stated pilot implementation.
D:Preventive posture work addresses overexposed data before a malicious AI interaction occurs. This describes the sound approach but does not diagnose why the observed alternative failed.