Microsoft Updated AB-620 Exam Questions and Answers by cataleya

A. Configure prompt modifications at the system level. Centralized instructions enforce the required tone, disclaimers, response boundaries, and refusal behavior consistently across generative responses. Microsoft documents that prompt modification can define an agent’s personality, response format, permitted subjects, and behavior when an answer is unavailable. This avoids inconsistent manual changes across individual topics. Microsoft: Prompt modification

D. Configure Power Platform DLP policies. DLP policies govern which connectors, knowledge sources, HTTP endpoints, and channels the agent may use. They should be configured and validated before publication because a policy violation can prevent the agent from being published. DLP also protects against unauthorized data movement between business and non-business connectors. Microsoft: Data policies for agents

Why the others are incorrect: B creates duplicated, inconsistent disclaimers. C incorrectly treats moderation as a web-channel setting; moderation is configured for generative behavior rather than enforced only by embedding settings. E applies safeguards after publication and assumes filters must be configured separately for every channel, which would not provide the required centralized governance.

Use the user ' s message → Select Activity.Text as the input; Query only designated data → Turn on Search only selected sources and choose the approved sources; Restrict to enterprise data → Turn off Allow the AI to use its own general knowledge.

Comprehensive and Detailed Explanation From Microsoft AB-620 Study Guide: Activity.Text is the system variable containing the current user ' s utterance, so it is the appropriate query input for the generative answers node. Search only selected sources changes the node from the default broad agent-level search to an explicit source set for that topic. Turning off the option that permits the model to use its own general knowledge closes the ungrounded fallback path and ensures the response depends on approved enterprise content. Generative orchestration at the agent level does not itself guarantee that this particular node searches only designated sources. The selected sources replace, rather than supplement, the agent-level list for the node, and there is no automatic fallback to other knowledge if they return no answer. That behavior is desirable for strict policy boundaries but should be reflected in a clear no-answer or escalation response. Testing should inspect source use, verify authorization trimming under different users, and include questions for which the approved source contains no relevant material to ensure the agent does not invent an answer. Study Guide alignment: Plan and configure agent solutions > Configure topics > Configure generative answers node.

===============

Comprehensive and Detailed Explanation From Microsoft AB-620 Study Guide: Power Platform data policies are the preventive governance control for connectors, HTTP endpoints, knowledge sources, and combinations of data groups used by Copilot Studio agents. Administrators can classify or block connectors, restrict endpoint use, require authentication, and surface policy violations during authoring or publishing. This meets both enforcement and reporting requirements. Application Insights records runtime telemetry but does not prevent a maker from adding a disallowed connector. Security FAQs provide guidance only. Agent analytics can reveal behavior and usage trends but are detective, not an administrative guardrail. Data policies should be scoped deliberately at tenant or environment level and tested against representative agent configurations. When a violation occurs, Copilot Studio can block publishing and provide downloadable details identifying the policy conflict, enabling the administrator and maker to correct the design. Governance should also include least-privilege connections, environment segmentation, solution-based ALM, and review of exceptions; a broadly permissive policy followed by manual monitoring is not equivalent to enforcing the control at design and deployment time. Study Guide alignment: Plan and configure agent solutions > Plan an agent solution > Evaluate security and governance considerations.

===============

Comprehensive and Detailed Explanation From Microsoft AB-620 Study Guide: The first configuration check is the contract between the agent and the flow. Input and output parameter definitions must agree on name, data type, required status, and mapping. A flow can complete successfully yet return an empty value to the agent if the response field is not mapped to the declared output or if the agent expects a different type. Validating the parameter definitions therefore addresses the actual data-exchange requirement. Reviewing run history is useful for diagnosis, but it does not correct a mismatched contract by itself. Changing the trigger schema or adding Parse JSON might be necessary in a specific implementation, yet neither should be done before confirming what the caller sends and what the flow returns. Republishing cannot repair an incorrect mapping. The builder should test the flow directly with known values, inspect each action output, verify the final response action, and then test from the agent. Optional and null fields should be handled explicitly, and sensitive outputs should be minimized and protected in logs. Study Guide alignment: Plan and configure agent solutions > Create and monitor agent flows in Copilot Studio > Add input and output parameters.

===============