McAfee Secure

Microsoft AB-900 Bundle

Exam Code: AB-900

Exam Name Microsoft 365 Copilot and Agent Administration Fundamentals

Certification Provider: Microsoft

Corresponding Certification: Microsoft 365 Certified: Copilot and Agent Administration Fundamentals

Microsoft AB-900 Bundle $19.99

100% Microsoft AB-900 Guaranteed Success With Testking Training Materials

Get Real AB-900 Exam Questions & Expert Verified Answers as Experienced on the Real Exam!

  • Questions & Answers

    AB-900 Questions & Answers

    99 Questions & Answers

    The ultimate exam preparation tool, Questions & Answers mirrors the actual AB-900 exam while allowing you to hone and measure every aspect of your test-taking ability.

  • Study Guide

    AB-900 Study Guide

    314 PDF Pages

    Developed by industry experts, this 314-page guide spells out in painstaking detail all of the information you need to ace AB-900 exam.

Microsoft 365 Copilot and Agent Administration Fundamentals (AB-900) Practice Tests

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
Total Cost: $154.98
Bundle Price: $134.99

Purchase Individually

  • Questions & Answers

    Questions & Answers

    99 Questions

    $124.99
  • Study Guide

    Study Guide

    314 PDF Pages

    $29.99

Microsoft AB-900 Practice Test 20

Question 190

A Microsoft 365 administrator implemented the following change: Require human review and source validation for consequential AI-generated content. The business requirement was: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. Which observed test result would best confirm that the change worked?

  1. The removed user cannot open the protected content and eventual search results reflect the new access boundary.
  2. The user can view content only when their label-derived rights allow it.
  3. The proposed advice is checked against verified evidence and revised before release.
  4. The coworker sees the shared response but cannot open the protected document unless independent authorized SharePoint access is granted.

Explanation

Correct Answer

C:

Responsible AI includes accountability, fairness, transparency and oversight beyond mere model output generation. A saved configuration without effective access or behavior evidence would be insufficient.

Incorrect Answers

A:

The removed user cannot open the protected content and eventual search results reflect the new access boundary. However, this does not establish the required outcome: the proposed advice is checked against verified evidence and revised before release.

B:

The user can view content only when their label-derived rights allow it. However, this does not establish the required outcome: the proposed advice is checked against verified evidence and revised before release.

D:

The coworker sees the shared response but cannot open the protected document unless independent authorized SharePoint access is granted. However, this does not establish the required outcome: the proposed advice is checked against verified evidence and revised before release.

Question 191

For the stated requirement, an administrator tried this control: Require human review and source validation for consequential AI-generated content. The expected need was: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. However, the result was: Generated claims are sent to customers with no validation of factual or fairness risks. Which cause best explains this specific failure?

  1. The role of Microsoft Graph grounding and authorized service context was overlooked.
  2. A human validation and responsible-AI review gate was omitted.
  3. The policy location or supported interaction scope was not included.
  4. The matching detection definitions or scope do not include the identifiers and repository.

Explanation

Correct Answer

B:

Responsible AI includes accountability, fairness, transparency and oversight beyond mere model output generation. A repair should return to the action: Require human review and source validation for consequential AI-generated content.

Incorrect Answers

A:

The role of Microsoft Graph grounding and authorized service context was overlooked. The stated evidence instead indicates: A human validation and responsible-AI review gate was omitted.

C:

The policy location or supported interaction scope was not included. The stated evidence instead indicates: A human validation and responsible-AI review gate was omitted.

D:

The matching detection definitions or scope do not include the identifiers and repository. The stated evidence instead indicates: A human validation and responsible-AI review gate was omitted.

Question 192

An IT team has TWO independent requirements. Requirement A: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. Requirement B: A manager reports possible offensive language in a Teams exchange and needs the compliance reviewer to inspect policy matches. Which TWO actions together satisfy these separate requirements? (Select two.)

  1. Review the relevant Microsoft Purview data security posture and AI observability views.
  2. Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI.
  3. Require human review and source validation for consequential AI-generated content.
  4. Use the Communication Compliance reviewer workflow and inspect relevant messages under the policy.
  5. Use SharePoint restricted content discovery where supported for the selected site.

Explanation

Correct Answers

C:

This is the necessary action for requirement A. Responsible AI includes accountability, fairness, transparency and oversight beyond mere model output generation. Verify A by checking: The proposed advice is checked against verified evidence and revised before release.

D:

This is the necessary action for requirement B. Communication Compliance review addresses communications policy violations, not generic sensitive-file alerts. Verify B by checking: The review queue contains the matching communication and reviewer decision history.

Incorrect Answers

A:

“Review the relevant Microsoft Purview data security posture and AI observability views” addresses another source-defined requirement: An administrator is evaluating this Microsoft 365 requirement: A security office wants a consolidated picture of where sensitive organizational data is exposed in AI-related activity before selecting mitigations. Its controlling reason is: Data security posture management for AI offers risk visibility; DLP incident response is a separate control. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. Requirement B: A manager reports possible offensive language in a Teams exchange and needs the compliance reviewer to inspect policy matches. Both original keyed selections are required: “Require human review and source validation for consequential AI-generated content” and “Use the Communication Compliance reviewer workflow and inspect relevant messages under the policy”; this alternative does not replace either selected requirement.

B:

“Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question.…. Its controlling reason is: DSPM for AI supports AI-related activity visibility and data-risk discovery, not automatic access revocation. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. Requirement B: A manager reports possible offensive language in a Teams exchange and needs the compliance reviewer to inspect policy matches. Both original keyed selections are required: “Require human review and source validation for consequential AI-generated content” and “Use the Communication Compliance reviewer workflow and inspect relevant messages under the policy”; this alternative does not replace either selected requirement.

E:

“Use SharePoint restricted content discovery where supported for the selected site” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: A compliance manager wants to prioritize regulatory control improvements and track their completion rather than investigate one suspicious file. Requirement B:…. Its controlling reason is: Restricted content discovery addresses search surfacing and is not equivalent to restricted access control. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A marketing team wants AI-generated client advice to be reviewed for errors and unfair assumptions before sending it externally. Requirement B: A manager reports possible offensive language in a Teams exchange and needs the compliance reviewer to inspect policy matches. Both original keyed selections are required: “Require human review and source validation for consequential AI-generated content” and “Use the Communication Compliance reviewer workflow and inspect relevant messages under the policy”; this alternative does not replace either selected requirement.

Question 193

An administrator is evaluating this Microsoft 365 requirement: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Which action best meets the specified requirement?

  1. Check the effective sensitivity-label encryption usage rights before assuming the Copilot service malfunctioned.
  2. Apply supported Purview data protection controls and review the specific Copilot interaction protection coverage.
  3. Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context.
  4. Validate the effective site permission change and allow for supported search index propagation before concluding access persists.

Explanation

Correct Answer

D:

Search propagation and actual content authorization are separate signals and should be tested independently. The acceptance check is: The removed user cannot open the protected content and eventual search results reflect the new access boundary.

Incorrect Answers

A:

“Check the effective sensitivity-label encryption usage rights before assuming the Copilot service malfunctioned” responds to another technical or business failure: An IT team has TWO independent requirements. Requirement A: A company needs confidential board files to carry a classification and encryption requirements when employees download or email them. Requirement B: An…. That source case depends on: A discoverable link is not proof the user has the cryptographic rights needed to open content. The present issue is: An administrator is evaluating this Microsoft 365 requirement: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. The more direct action is “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” because Search propagation and actual content authorization are separate signals and should be tested independently.

B:

“Apply supported Purview data protection controls and review the specific Copilot interaction protection coverage” responds to another technical or business failure: An IT team has TWO independent requirements. Requirement A: A hospital must recognize patient identifiers in many documents before deciding which protection policies to apply. Requirement B: A tenant wants to prevent…. That source case depends on: Copilot interactions can be subject to supported data governance controls but not every rule applies identically to every AI surface. The present issue is: An administrator is evaluating this Microsoft 365 requirement: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. The more direct action is “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” because Search propagation and actual content authorization are separate signals and should be tested independently.

C:

“Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context” responds to another technical or business failure: An administrator is evaluating this Microsoft 365 requirement: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. That source case depends on: Grounding can enrich responses from authorized organizational content; it does not erase access controls. The present issue is: An administrator is evaluating this Microsoft 365 requirement: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. The more direct action is “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” because Search propagation and actual content authorization are separate signals and should be tested independently.

Question 194

A Microsoft 365 administrator implemented the following change: Validate the effective site permission change and allow for supported search index propagation before concluding access persists. The business requirement was: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Which observed test result would best confirm that the change worked?

  1. The user can view content only when their label-derived rights allow it.
  2. The response's source references map to data accessible to the requester, not unrelated tenant secrets.
  3. The removed user cannot open the protected content and eventual search results reflect the new access boundary.
  4. A controlled sensitive interaction follows the applicable protection and logging behavior.

Explanation

Correct Answer

C:

Search propagation and actual content authorization are separate signals and should be tested independently. A saved configuration without effective access or behavior evidence would be insufficient.

Incorrect Answers

A:

The user can view content only when their label-derived rights allow it. However, this does not establish the required outcome: the removed user cannot open the protected content and eventual search results reflect the new access boundary.

B:

The response's source references map to data accessible to the requester, not unrelated tenant secrets. However, this does not establish the required outcome: the removed user cannot open the protected content and eventual search results reflect the new access boundary.

D:

A controlled sensitive interaction follows the applicable protection and logging behavior. However, this does not establish the required outcome: the removed user cannot open the protected content and eventual search results reflect the new access boundary.

Question 195

For the stated requirement, an administrator tried this control: Validate the effective site permission change and allow for supported search index propagation before concluding access persists. The expected need was: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. However, the result was: The user can still directly open the site after the purported removal. Which cause best explains this specific failure?

  1. The policy location or supported interaction scope was not included.
  2. Information Protection encryption usage rights exclude the user.
  3. The user's effective access remains through another site role or group rather than a search-only delay.
  4. The role of Microsoft Graph grounding and authorized service context was overlooked.

Explanation

Correct Answer

C:

Search propagation and actual content authorization are separate signals and should be tested independently. A repair should return to the action: Validate the effective site permission change and allow for supported search index propagation before concluding access persists.

Incorrect Answers

A:

The policy location or supported interaction scope was not included. The stated evidence instead indicates: The user's effective access remains through another site role or group rather than a search-only delay.

B:

Information Protection encryption usage rights exclude the user. The stated evidence instead indicates: The user's effective access remains through another site role or group rather than a search-only delay.

D:

The role of Microsoft Graph grounding and authorized service context was overlooked. The stated evidence instead indicates: The user's effective access remains through another site role or group rather than a search-only delay.

Question 196

An IT team has TWO independent requirements. Requirement A: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Requirement B: An administrator needs a timeline of data-label activity to validate whether a sensitivity label was removed from a document. Which TWO actions together satisfy these separate requirements? (Select two.)

  1. Run SharePoint data access governance reporting and examine sites with broad effective access.
  2. Use SharePoint restricted content discovery where supported for the selected site.
  3. Validate the effective site permission change and allow for supported search index propagation before concluding access persists.
  4. Complete the supported Communication Compliance review and escalation workflow for the matching item.
  5. Review relevant events in Microsoft Purview Activity Explorer with appropriate filters.

Explanation

Correct Answers

C:

This is the necessary action for requirement A. Search propagation and actual content authorization are separate signals and should be tested independently. Verify A by checking: The removed user cannot open the protected content and eventual search results reflect the new access boundary.

E:

This is the necessary action for requirement B. Activity Explorer reports supported user and policy activities over time; current labels show state. Verify B by checking: The timeline shows label-related activities matching the file and time window.

Incorrect Answers

A:

“Run SharePoint data access governance reporting and examine sites with broad effective access” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: A DLP alert indicates that a sales user tried to share customer identifiers externally; an analyst must determine the triggering action. Requirement B: A…. Its controlling reason is: Oversharing assessment starts with broad-access evidence rather than simply changing Copilot settings. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Requirement B: An administrator needs a timeline of data-label activity to validate whether a sensitivity label was removed from a document. Both original keyed selections are required: “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” and “Review relevant events in Microsoft Purview Activity Explorer with appropriate filters”; this alternative does not replace either selected requirement.

B:

“Use SharePoint restricted content discovery where supported for the selected site” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: A compliance manager wants to prioritize regulatory control improvements and track their completion rather than investigate one suspicious file. Requirement B:…. Its controlling reason is: Restricted content discovery addresses search surfacing and is not equivalent to restricted access control. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Requirement B: An administrator needs a timeline of data-label activity to validate whether a sensitivity label was removed from a document. Both original keyed selections are required: “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” and “Review relevant events in Microsoft Purview Activity Explorer with appropriate filters”; this alternative does not replace either selected requirement.

D:

“Complete the supported Communication Compliance review and escalation workflow for the matching item” addresses another source-defined requirement: An administrator is evaluating this Microsoft 365 requirement: A communication policy flags a message but the compliance team must demonstrate a documented reviewer decision and escalation. Its controlling reason is: A detected communication policy match is not the same as a completed compliance review. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: An employee was removed from a restricted site yesterday, but Copilot search still briefly shows stale search metadata. Requirement B: An administrator needs a timeline of data-label activity to validate whether a sensitivity label was removed from a document. Both original keyed selections are required: “Validate the effective site permission change and allow for supported search index propagation before concluding access persists” and “Review relevant events in Microsoft Purview Activity Explorer with appropriate filters”; this alternative does not replace either selected requirement.

Question 197

An administrator is evaluating this Microsoft 365 requirement: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Which action best meets the specified requirement?

  1. Apply supported Purview data protection controls and review the specific Copilot interaction protection coverage.
  2. Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context.
  3. Inspect the Microsoft Graph connector ingestion status, index freshness, and mapped access-control lists before changing the Copilot prompt.
  4. Review and correct the underlying SharePoint file permissions rather than assume Copilot bypassed them.

Explanation

Correct Answer

B:

Grounding can enrich responses from authorized organizational content; it does not erase access controls. The acceptance check is: The response's source references map to data accessible to the requester, not unrelated tenant secrets.

Incorrect Answers

A:

“Apply supported Purview data protection controls and review the specific Copilot interaction protection coverage” responds to another technical or business failure: An IT team has TWO independent requirements. Requirement A: A hospital must recognize patient identifiers in many documents before deciding which protection policies to apply. Requirement B: A tenant wants to prevent…. That source case depends on: Copilot interactions can be subject to supported data governance controls but not every rule applies identically to every AI surface. The present issue is: An administrator is evaluating this Microsoft 365 requirement: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. The more direct action is “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context” because Grounding can enrich responses from authorized organizational content; it does not erase access controls.

C:

“Inspect the Microsoft Graph connector ingestion status, index freshness, and mapped access-control lists before changing the Copilot prompt” responds to another technical or business failure: An IT team has TWO independent requirements. Requirement A: A records manager needs project emails preserved for a legally approved duration before eligible disposal. Requirement B: An authorized employee cannot find…. That source case depends on: Microsoft Graph connector content can ground Copilot only after ingestion and indexing that preserves the source permission boundary. The present issue is: An administrator is evaluating this Microsoft 365 requirement: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. The more direct action is “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context” because Grounding can enrich responses from authorized organizational content; it does not erase access controls.

D:

“Review and correct the underlying SharePoint file permissions rather than assume Copilot bypassed them” responds to another technical or business failure: An IT team has TWO independent requirements. Requirement A: A company must prevent employees from posting credit-card data in a targeted Teams chat while allowing normal text discussion. Requirement B: A user asks…. That source case depends on: Copilot grounding honors authorized data access; prompt wording is not a substitute for permissions remediation. The present issue is: An administrator is evaluating this Microsoft 365 requirement: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. The more direct action is “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context” because Grounding can enrich responses from authorized organizational content; it does not erase access controls.

Question 198

A Microsoft 365 administrator implemented the following change: Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context. The business requirement was: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Which observed test result would best confirm that the change worked?

  1. The response's source references map to data accessible to the requester, not unrelated tenant secrets.
  2. The file retains the label and authorized users can open it while an unauthorized account cannot decrypt its contents.
  3. The user can view content only when their label-derived rights allow it.
  4. After access is removed and changes propagate, the user cannot open or retrieve the protected file through authorized search and Copilot.

Explanation

Correct Answer

A:

Grounding can enrich responses from authorized organizational content; it does not erase access controls. A saved configuration without effective access or behavior evidence would be insufficient.

Incorrect Answers

B:

The file retains the label and authorized users can open it while an unauthorized account cannot decrypt its contents. However, this does not establish the required outcome: the response's source references map to data accessible to the requester, not unrelated tenant secrets.

C:

The user can view content only when their label-derived rights allow it. However, this does not establish the required outcome: the response's source references map to data accessible to the requester, not unrelated tenant secrets.

D:

After access is removed and changes propagate, the user cannot open or retrieve the protected file through authorized search and Copilot. However, this does not establish the required outcome: the response's source references map to data accessible to the requester, not unrelated tenant secrets.

Question 199

For the stated requirement, an administrator tried this control: Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context. The expected need was: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. However, the result was: A reviewer incorrectly assumes prompt text alone determines every grounded answer. Which cause best explains this specific failure?

  1. The DLP policy is not scoped to Teams or its detection rule does not match the information.
  2. The policy location or supported interaction scope was not included.
  3. A human validation and responsible-AI review gate was omitted.
  4. The role of Microsoft Graph grounding and authorized service context was overlooked.

Explanation

Correct Answer

D:

Grounding can enrich responses from authorized organizational content; it does not erase access controls. A repair should return to the action: Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context.

Incorrect Answers

A:

The DLP policy is not scoped to Teams or its detection rule does not match the information. The stated evidence instead indicates: The role of Microsoft Graph grounding and authorized service context was overlooked.

B:

The policy location or supported interaction scope was not included. The stated evidence instead indicates: The role of Microsoft Graph grounding and authorized service context was overlooked.

C:

A human validation and responsible-AI review gate was omitted. The stated evidence instead indicates: The role of Microsoft Graph grounding and authorized service context was overlooked.

Question 200

An IT team has TWO independent requirements. Requirement A: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Requirement B: A policy owner wants to discover which AI apps or agents are using organizational data before designing tighter safeguards. Which TWO actions together satisfy these separate requirements? (Select two.)

  1. Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI.
  2. Explore supported Microsoft Purview Data Explorer information about classified sensitive items.
  3. Use Microsoft Purview Compliance Manager assessments and improvement actions.
  4. Use appropriate Microsoft Purview sensitive information types or classifiers to discover matching data.
  5. Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context.

Explanation

Correct Answers

A:

This is the necessary action for requirement B. DSPM for AI supports AI-related activity visibility and data-risk discovery, not automatic access revocation. Verify B by checking: The identified AI applications and activity patterns support an evidence-based safeguard plan.

E:

This is the necessary action for requirement A. Grounding can enrich responses from authorized organizational content; it does not erase access controls. Verify A by checking: The response's source references map to data accessible to the requester, not unrelated tenant secrets.

Incorrect Answers

B:

“Explore supported Microsoft Purview Data Explorer information about classified sensitive items” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: An encrypted file has a confidentiality label and a user receives the file link in Copilot but cannot open the content. Requirement B: A privacy officer wants…. Its controlling reason is: Data Explorer helps locate sensitive-content distribution, not declare a policy violation by itself. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Requirement B: A policy owner wants to discover which AI apps or agents are using organizational data before designing tighter safeguards. Both original keyed selections are required: “Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI” and “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context”; this alternative does not replace either selected requirement.

C:

“Use Microsoft Purview Compliance Manager assessments and improvement actions” addresses another source-defined requirement: An IT team has TWO independent requirements. Requirement A: An authorized employee cannot find recent approved knowledge pages through Copilot after an enterprise wiki was added by a Microsoft Graph connector.…. Its controlling reason is: Compliance Manager evaluates control posture and improvement activity; it does not replace all operational investigations. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Requirement B: A policy owner wants to discover which AI apps or agents are using organizational data before designing tighter safeguards. Both original keyed selections are required: “Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI” and “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context”; this alternative does not replace either selected requirement.

D:

“Use appropriate Microsoft Purview sensitive information types or classifiers to discover matching data” addresses another source-defined requirement: An administrator is evaluating this Microsoft 365 requirement: A hospital must recognize patient identifiers in many documents before deciding which protection policies to apply. Its controlling reason is: Data classification depends on the correct detector and scanning scope, not only a policy name. This question asks for two independent business requirements: An IT team has TWO independent requirements. Requirement A: A manager needs to explain why Copilot responses mention information from several permitted Teams and SharePoint sources instead of only the typed question. Requirement B: A policy owner wants to discover which AI apps or agents are using organizational data before designing tighter safeguards. Both original keyed selections are required: “Review AI activity and risk insights in Microsoft Purview Data Security Posture Management for AI” and “Explain the Microsoft Graph grounding process using data accessible in the signed-in user's context”; this alternative does not replace either selected requirement.