Exam Bundle

Approved AI usage meets policy without transmitting protected information. This would be relevant evidence for a different failure: A security team bans every AI destination rather than applying risk-based safeguards for approved work, usually addressed by Define AI use policies with destination trust, content controls, and monitoring appropriate to risk. It does not demonstrate that the current problem at Aster Mutual was resolved. The specific required evidence is: A protected test document triggers the expected DLP event and a non-sensitive file passes.

An exfiltration simulation fails and produces a reviewable security event. This would be relevant evidence for a different failure: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload, usually addressed by Enforce trusted authorization and AI guardrails together with outbound data inspection. It does not demonstrate that the current problem at Aster Mutual was resolved. The specific required evidence is: A protected test document triggers the expected DLP event and a non-sensitive file passes.

A prohibited data transfer is stopped even when the destination is otherwise allowed. This would be relevant evidence for a different failure: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites, usually addressed by Narrow exceptions to specific users, destinations, and justified content classifications. It does not demonstrate that the current problem at Aster Mutual was resolved. The specific required evidence is: A protected test document triggers the expected DLP event and a non-sensitive file passes.

Apply data loss prevention classification and content controls for upload actions. That action fits another problem: An organization allows file uploads to sanctioned SaaS but has no inspection for sensitive document classification. It does not directly remediate the current finding at Horizon Energy Services: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

Configure DLP policy for sensitive identifiers on governed web and AI destinations. That action fits another problem: Employees paste confidential customer numbers into public generative AI sites from company browsers. It does not directly remediate the current finding at Horizon Energy Services: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

Narrow exceptions to specific users, destinations, and justified content classifications. That action fits another problem: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. It does not directly remediate the current finding at Horizon Energy Services: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

A protected test document triggers the expected DLP event and a non-sensitive file passes. This would be relevant evidence for a different failure: An organization allows file uploads to sanctioned SaaS but has no inspection for sensitive document classification, usually addressed by Apply data loss prevention classification and content controls for upload actions. It does not demonstrate that the current problem at Horizon Energy Services was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

A prohibited data transfer is stopped even when the destination is otherwise allowed. This would be relevant evidence for a different failure: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites, usually addressed by Narrow exceptions to specific users, destinations, and justified content classifications. It does not demonstrate that the current problem at Horizon Energy Services was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

A simulated restricted upload is blocked or reviewed while safe requests remain usable. This would be relevant evidence for a different failure: Employees paste confidential customer numbers into public generative AI sites from company browsers, usually addressed by Configure DLP policy for sensitive identifiers on governed web and AI destinations. It does not demonstrate that the current problem at Horizon Energy Services was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. That action fits another problem: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched. It does not directly remediate the current finding at Newbridge Charity: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

Narrow exceptions to specific users, destinations, and justified content classifications. That action fits another problem: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. It does not directly remediate the current finding at Newbridge Charity: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

Enforce trusted authorization and AI guardrails together with outbound data inspection. That action fits another problem: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. It does not directly remediate the current finding at Newbridge Charity: A security team bans every AI destination rather than applying risk-based safeguards for approved work. The relevant corrective control for this case is: Define AI use policies with destination trust, content controls, and monitoring appropriate to risk.

An exfiltration simulation fails and produces a reviewable security event. This would be relevant evidence for a different failure: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload, usually addressed by Enforce trusted authorization and AI guardrails together with outbound data inspection. It does not demonstrate that the current problem at Newbridge Charity was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

Analysts can justify an incident from safe metadata and approved uploads are no longer blocked. This would be relevant evidence for a different failure: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched, usually addressed by Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. It does not demonstrate that the current problem at Newbridge Charity was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

A prohibited data transfer is stopped even when the destination is otherwise allowed. This would be relevant evidence for a different failure: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites, usually addressed by Narrow exceptions to specific users, destinations, and justified content classifications. It does not demonstrate that the current problem at Newbridge Charity was resolved. The specific required evidence is: Approved AI usage meets policy without transmitting protected information.

Apply data loss prevention classification and content controls for upload actions. That action fits another problem: An organization allows file uploads to sanctioned SaaS but has no inspection for sensitive document classification. It does not directly remediate the current finding at Harbor Point Medical: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

Define AI use policies with destination trust, content controls, and monitoring appropriate to risk. That action fits another problem: A security team bans every AI destination rather than applying risk-based safeguards for approved work. It does not directly remediate the current finding at Harbor Point Medical: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

Enforce trusted authorization and AI guardrails together with outbound data inspection. That action fits another problem: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. It does not directly remediate the current finding at Harbor Point Medical: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

An exfiltration simulation fails and produces a reviewable security event. This would be relevant evidence for a different failure: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload, usually addressed by Enforce trusted authorization and AI guardrails together with outbound data inspection. It does not demonstrate that the current problem at Harbor Point Medical was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

A protected test document triggers the expected DLP event and a non-sensitive file passes. This would be relevant evidence for a different failure: An organization allows file uploads to sanctioned SaaS but has no inspection for sensitive document classification, usually addressed by Apply data loss prevention classification and content controls for upload actions. It does not demonstrate that the current problem at Harbor Point Medical was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

Approved AI usage meets policy without transmitting protected information. This would be relevant evidence for a different failure: A security team bans every AI destination rather than applying risk-based safeguards for approved work, usually addressed by Define AI use policies with destination trust, content controls, and monitoring appropriate to risk. It does not demonstrate that the current problem at Harbor Point Medical was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

Configure DLP policy for sensitive identifiers on governed web and AI destinations. That action fits another problem: Employees paste confidential customer numbers into public generative AI sites from company browsers. It does not directly remediate the current finding at Clearbrook University: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. That action fits another problem: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched. It does not directly remediate the current finding at Clearbrook University: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

Enforce trusted authorization and AI guardrails together with outbound data inspection. That action fits another problem: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. It does not directly remediate the current finding at Clearbrook University: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. The relevant corrective control for this case is: Narrow exceptions to specific users, destinations, and justified content classifications.

Analysts can justify an incident from safe metadata and approved uploads are no longer blocked. This would be relevant evidence for a different failure: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched, usually addressed by Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. It does not demonstrate that the current problem at Clearbrook University was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

A simulated restricted upload is blocked or reviewed while safe requests remain usable. This would be relevant evidence for a different failure: Employees paste confidential customer numbers into public generative AI sites from company browsers, usually addressed by Configure DLP policy for sensitive identifiers on governed web and AI destinations. It does not demonstrate that the current problem at Clearbrook University was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

An exfiltration simulation fails and produces a reviewable security event. This would be relevant evidence for a different failure: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload, usually addressed by Enforce trusted authorization and AI guardrails together with outbound data inspection. It does not demonstrate that the current problem at Clearbrook University was resolved. The specific required evidence is: A prohibited data transfer is stopped even when the destination is otherwise allowed.

Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. That action fits another problem: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched. It does not directly remediate the current finding at Oceanview Payments: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. The relevant corrective control for this case is: Enforce trusted authorization and AI guardrails together with outbound data inspection.

Narrow exceptions to specific users, destinations, and justified content classifications. That action fits another problem: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites. It does not directly remediate the current finding at Oceanview Payments: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. The relevant corrective control for this case is: Enforce trusted authorization and AI guardrails together with outbound data inspection.

Define AI use policies with destination trust, content controls, and monitoring appropriate to risk. That action fits another problem: A security team bans every AI destination rather than applying risk-based safeguards for approved work. It does not directly remediate the current finding at Oceanview Payments: A third-party prompt asks an enterprise AI assistant to disclose internal secrets through browser upload. The relevant corrective control for this case is: Enforce trusted authorization and AI guardrails together with outbound data inspection.

Analysts can justify an incident from safe metadata and approved uploads are no longer blocked. This would be relevant evidence for a different failure: DLP blocks legitimate encrypted uploads without showing analysts which sensitive pattern matched, usually addressed by Tune policy classification, severity, and event visibility without exposing protected content unnecessarily. It does not demonstrate that the current problem at Oceanview Payments was resolved. The specific required evidence is: An exfiltration simulation fails and produces a reviewable security event.

Approved AI usage meets policy without transmitting protected information. This would be relevant evidence for a different failure: A security team bans every AI destination rather than applying risk-based safeguards for approved work, usually addressed by Define AI use policies with destination trust, content controls, and monitoring appropriate to risk. It does not demonstrate that the current problem at Oceanview Payments was resolved. The specific required evidence is: An exfiltration simulation fails and produces a reviewable security event.

A prohibited data transfer is stopped even when the destination is otherwise allowed. This would be relevant evidence for a different failure: A secure Internet gateway has DLP rules but exceptions are so broad that personal information leaves through approved sites, usually addressed by Narrow exceptions to specific users, destinations, and justified content classifications. It does not demonstrate that the current problem at Oceanview Payments was resolved. The specific required evidence is: An exfiltration simulation fails and produces a reviewable security event.