Pass Your Microsoft AB-900 Exam Easy!

Privacy principles reduce unnecessary processing and disclosure. This helps with protection of sensitive personal information instead of the required verified high-stakes summary accuracy.

Accountability identifies responsible humans for monitoring and correction. The action is relevant to clear ownership of AI failures rather than verified high-stakes summary accuracy.

Accountability retains control of high-impact actions with authorized decision-makers. This helps with human oversight of consequential action instead of the required verified high-stakes summary accuracy.

Prompt wording alone cannot secure a powerful tool connection. That addresses enforceable permission checks for agent actions, whereas the current case requires safe handling of adversarial document text.

Monitoring enables investigation of attempts to misuse the assistant. It is used for evidence of AI abuse attempts; this scenario calls for safe handling of adversarial document text.

Source text must not override trusted policy or backend authorization. That addresses resistance to malicious source instructions, whereas the current case requires safe handling of adversarial document text.

A Copilot summary does not remove legal preservation requirements on underlying content. It would support ongoing legal preservation of source evidence, not the stated mandated AI conversation retention.

Content lifecycle operates on source records independent of Copilot summaries. It would support policy-based deletion of expired source content, not the stated mandated AI conversation retention.

Case-based content search preserves and locates relevant AI records. Its purpose is discovery of AI interaction evidence; the specified need is mandated AI conversation retention.

Grounding uncertainty prevents fabricated specific claims. The action is relevant to safe answer when evidence is absent rather than a current authoritative document reference.

Static documents may not represent rapidly changing transactional state. This helps with current operational data rather than stale export instead of the required a current authoritative document reference.

Citations allow checking statements against evidence. This helps with traceable grounded policy statement instead of the required a current authoritative document reference.

Connector refresh keeps indexed items aligned with source changes. That addresses up-to-date external search index, whereas the current case requires corrected external item authorization.

Runtime operations are distinct from search grounding over indexed material. It is used for live action against a business system; this scenario calls for corrected external item authorization.

Indexed external items must carry permissions so results can be security trimmed. That addresses secure retrieval of connected external content, whereas the current case requires corrected external item authorization.

AI security telemetry can identify supported sensitive data interaction patterns. It would support detection of sensitive prompt content, not the stated source-document classification evidence.

Controls on the underlying source govern what data a user can access. Its purpose is consistent source access enforcement; the specified need is source-document classification evidence.

DLP can govern sensitive data handling in supported interactions and endpoints. It would support restriction of classified generated content movement, not the stated source-document classification evidence.

Correcting excessive authorization closes both direct and Copilot discovery paths. This helps with removal of internal oversharing instead of the required persistent financial-file control.

Data loss prevention provides incident context and prevention actions. This helps with outbound sensitive-mail incident response instead of the required persistent financial-file control.

DSPM for AI consolidates AI exposure findings and control opportunities. The action is relevant to AI-related data exposure reduction rather than persistent financial-file control.

Human control mitigates high-impact automation risks. It is used for accountable irreversible account deletion; this scenario calls for informed AI output interpretation.

Data minimization limits personal data processing for the task. That addresses least-data disclosure by an AI tool, whereas the current case requires informed AI output interpretation.

Purpose limitation and governance constrain sensitive processing. That addresses accountable processing of customer information, whereas the current case requires informed AI output interpretation.

Anonymous links may create an unintended access path beyond tenant membership. It would support closure of public sharing exposure, not the stated diagnosing absent authorized source.

AI activity telemetry records supported user interactions and classifications. Its purpose is observational AI usage evidence; the specified need is diagnosing absent authorized source.

Access control defects are not fixed by changing a prompt alone. It would support stopping unintended broad directory access, not the stated diagnosing absent authorized source.

Security trimming and individual context yield different retrieved evidence. This helps with explaining user-specific Copilot answers instead of the required eliminating an inherited oversharing root cause.

Copilot respects effective user access rather than granting new file rights. The action is relevant to denying retrieval of an inaccessible source rather than eliminating an inherited oversharing root cause.

Graph uses accessible work context when building grounded responses. This helps with summarizing accessible mailbox content instead of the required eliminating an inherited oversharing root cause.

The relevant access risk comes from content sharing and identity authorization. That addresses preventing unapproved contractor discovery, whereas the current case requires connector-grounded security trimming.

Grounding checks help distinguish actual retrieved facts from generated claims. That addresses verification against authorized evidence, whereas the current case requires connector-grounded security trimming.

Microsoft 365 Copilot accesses work content under the user's entitled resource scope. It is used for ordinary user-scoped work grounding; this scenario calls for connector-grounded security trimming.