A is incorrect: Copilot does not ignore selected text. When you select content and invoke Copilot, it uses that specific text as the basis for rewriting. The selected content is the primary context that Copilot transforms according to your instructions. Reference: https://support.microsoft.com/en-us/office/rewrite-text-with-copilot-in-word-923d9763-f896-4da7-8a3f-5b12c3bfc475

B is correct: When you select text in Word, Copilot uses that selected content as context for your instruction. It generates rewritten alternatives that you can navigate through and preview. You then choose to replace the original, insert below, or discard the result. Reference: https://support.microsoft.com/en-us/office/rewrite-text-with-copilot-in-word-923d9763-f896-4da7-8a3f-5b12c3bfc475

C is incorrect: Copilot processes text within the Microsoft 365 environment using its AI model and typically generates multiple rewritten versions. It considers your instruction and the selected text context. It does not send content to a separate external processing service. Reference: https://support.microsoft.com/en-us/office/rewrite-text-with-copilot-in-word-923d9763-f896-4da7-8a3f-5b12c3bfc475

D is incorrect: You do not need to copy text into the chat pane. Copilot's in-line rewrite and Draft on Selection features work directly on highlighted text within the document. Select the text, choose the Copilot icon, and provide your instruction. Reference: https://support.microsoft.com/en-us/office/rewrite-text-with-copilot-in-word-923d9763-f896-4da7-8a3f-5b12c3bfc475

A is incorrect: Keeping the Copilot window open is impractical and does not provide a proactive alert. The email notification option in the scheduling settings is designed to alert you when the response is ready without requiring you to monitor the window. Reference: https://support.microsoft.com/en-us/topic/schedule-copilot-prompts-29dfd5fb-211a-4515-88a6-730b8074e489

B is correct: When scheduling a prompt, one of the configurable delivery settings is whether to receive an email notification when the response is ready. Enabling this option ensures you are promptly alerted and can review the output before your meeting. Reference: https://support.microsoft.com/en-us/topic/schedule-copilot-prompts-29dfd5fb-211a-4515-88a6-730b8074e489

C is incorrect: A calendar reminder is a manual workaround that does not confirm the response is actually ready. The built-in email notification option within the scheduled prompt settings is the correct feature for receiving an alert upon completion. Reference: https://support.microsoft.com/en-us/topic/schedule-copilot-prompts-29dfd5fb-211a-4515-88a6-730b8074e489

D is incorrect: There is no admin-configured Teams notification channel for scheduled prompt completions. The email notification toggle is a per-user setting you enable when scheduling each individual prompt within the Copilot Chat interface. Reference: https://support.microsoft.com/en-us/topic/schedule-copilot-prompts-29dfd5fb-211a-4515-88a6-730b8074e489

A is incorrect: Retyping a manual prompt repeatedly is inefficient for a recurring task. Adding the ready-made agent from the Agent Store provides a consistent, reusable solution. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/copilot-agent-store

B is incorrect: Rebuilding an agent that already exists adds unnecessary build and maintenance effort. When a published agent meets the need, adopt it directly from the Agent Store. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/copilot-agent-store

C is incorrect: Commissioning a custom build is unwarranted when a suitable prebuilt agent already exists in the Agent Store with no development required. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/copilot-agent-store

D is correct: Prebuilt agents from Microsoft and trusted partners are available in the Agent Store and require no development work. When one already meets the requirement, adding it is the fastest, lowest-effort choice. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/copilot-agent-store

A is incorrect: Copilot does not require a video recording. Transcription of spoken content is the key requirement for post-meeting Copilot access. Recording is recommended for the full recap experience, but the transcript is what enables Copilot to provide post-meeting insights. Reference: https://support.microsoft.com/en-us/office/use-copilot-in-microsoft-teams-meetings-0bf9dd3c-96f7-44e2-8bb8-790bedf066b1

B is incorrect: The scheduling method does not determine post-meeting Copilot availability. Whether the meeting was scheduled via Teams or Outlook, Copilot requires live transcription to be enabled for post-meeting access in the Recap tab. Reference: https://support.microsoft.com/en-us/office/use-copilot-in-microsoft-teams-meetings-0bf9dd3c-96f7-44e2-8bb8-790bedf066b1

C is correct: Transcription is not required to turn on Copilot during a meeting, but to prompt Copilot or see its conversation history after the meeting ends, live transcription must be enabled. Without a saved transcript, Copilot cannot generate post-meeting recaps or answer questions about the discussion. Reference: https://support.microsoft.com/en-us/office/use-copilot-in-microsoft-teams-meetings-0bf9dd3c-96f7-44e2-8bb8-790bedf066b1

D is incorrect: Copilot does not have a minimum attendee requirement. Post-meeting Copilot availability depends on whether live transcription was enabled during the meeting, not on the number of participants who attended the session. Reference: https://support.microsoft.com/en-us/office/use-copilot-in-microsoft-teams-meetings-0bf9dd3c-96f7-44e2-8bb8-790bedf066b1

A is incorrect: License type does not affect the visibility of saved prompts. Saved prompts are private by default regardless of which license other users hold. Sharing requires an explicit action through the Prompt Gallery. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/copilot-prompt-gallery

B is incorrect: Saved prompts are not automatically visible to the entire tenant. They are stored privately under Your prompts and are accessible only to the user who saved them. You must explicitly share them to make them available to others. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/copilot-prompt-gallery

C is correct: When you save a prompt in the Copilot Prompt Gallery, it appears in the Your prompts tab and is visible only to you. To make prompts available to others, you must take an explicit sharing action such as sharing to a Microsoft Teams team or copying the prompt link. Reference: https://support.microsoft.com/en-us/topic/how-to-save-prompts-55373730-2627-46a5-b0d8-772abe22dba4

D is incorrect: Managers do not have default access to view your saved prompts in any portal. Saved prompts in the Copilot Prompt Gallery are private to you by default. To make prompts available to others, you must explicitly share them. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/copilot-prompt-gallery-export-prompts

A is correct: A recurring need for consistent, scoped answers is well served by a custom agent. Configuring instructions defines how the agent should respond and limits it to your procedures, ensuring a uniform experience across the team. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/declarative-agent-best-practices

B is incorrect: Requiring everyone to craft detailed prompts each time is inconsistent and error-prone for a recurring need. A custom agent with configured instructions enforces consistent tone and scope automatically. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/agents-overview

C is incorrect: General chat responses do vary in tone and scope without configured instructions. A custom agent with instructions is the way to enforce consistency for a recurring team need. Reference: https://learn.microsoft.com/en-us/copilot/agents

D is incorrect: Funneling all questions through one person creates a bottleneck and does not scale. A custom agent with configured instructions lets everyone self-serve consistent, scoped answers. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/agent-essentials/m365-agents-admin-guide

A is incorrect: Copilot Chat does not automatically send email summaries when conversations end. Previous conversations are accessed through the Conversations list directly within Copilot Chat, not through Outlook emails. Reference: https://support.microsoft.com/en-us/topic/revisit-your-microsoft-365-copilot-chat-history-6ea899e3-3bb1-450a-a2ae-220341ac193a

B is correct: In Copilot Chat, the Conversations list displays your recent conversations. You can select any conversation to reopen and continue it, or select 'All conversations' to browse your entire chat history and find specific sessions from earlier. Reference: https://support.microsoft.com/en-us/topic/revisit-your-microsoft-365-copilot-chat-history-6ea899e3-3bb1-450a-a2ae-220341ac193a

C is incorrect: Copilot does not have a session recall feature that reproduces entire previous conversations on demand. To find a past conversation, navigate to the Conversations list in Copilot Chat and select the session directly. Reference: https://support.microsoft.com/en-us/topic/revisit-your-microsoft-365-copilot-chat-history-6ea899e3-3bb1-450a-a2ae-220341ac193a

D is incorrect: While the My Account portal allows you to manage and delete Copilot activity history, it is not designed for browsing or continuing individual conversations. Use the Conversations list directly in Copilot Chat to find and resume sessions. Reference: https://support.microsoft.com/en-us/office/delete-your-microsoft-365-copilot-activity-history-76de8afa-5eaf-43b0-bda8-0076d6e0390f

A is correct: A custom agent configured with the procedures library as its knowledge source ensures every query is answered from the same curated content, producing consistent and reliable results regardless of how members phrase their prompts. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/agent-builder-add-knowledge

B is incorrect: Personal copies create fragmentation and version drift, making answers less consistent. A single custom agent grounded in the shared SharePoint library keeps everyone aligned to one authoritative source. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/knowledge-sources

C is incorrect: Manual verification after every answer is inefficient and still allows inconsistent responses to surface. Grounding a custom agent in the SharePoint library is the proactive way to ensure source-aligned answers. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/agent-builder-build-agents

D is incorrect: Standardizing prompt wording does not control which content Copilot draws from, so answers can still vary. A custom agent grounded in the SharePoint library is what restricts responses to the authoritative source. Reference: https://learn.microsoft.com/en-us/microsoft-365/copilot/extensibility/knowledge-sources

A is correct: Agents respect existing permissions and security controls. When an agent invokes an action, the user must authenticate to the external system, and the external system's authorization rules determine what data is returned. Agents inherit all data protections provided by Microsoft 365 Copilot. Reference: https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/overview-declarative-agent

B is incorrect: Copilot does not automatically redact data returned from external systems. Data access is controlled through authentication and the external system's permissions model, not through post-retrieval redaction of all specific details. Reference: https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/agents-overview

C is incorrect: Agent actions do not bypass external system permissions. Users authenticate to external systems through the configured connection, and the external system enforces its own authorization rules alongside Microsoft 365 protections. Reference: https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/orchestrator

D is incorrect: Agent actions do not require per-query administrator approval. Security is managed through authenticated connections, the external system's access controls, and governance settings configured during agent deployment. Reference: https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/ecosystem

A is incorrect: Copilot does not aggregate data from multiple tenants. It is strictly scoped to one tenant at a time. A user must actively switch tenant context to access data in another tenant, and each response uses data from only the current tenant. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-privacy

B is incorrect: Guest access does not allow Copilot to combine data from multiple tenants in a single response. The user must switch tenant context, and data from each tenant is processed independently within that tenant's security boundary. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-privacy

C is correct: Logical isolation of customer content within each tenant is achieved through Microsoft Entra authorization and role-based access control. Copilot is strictly scoped to one tenant at a time and cannot query data across tenant boundaries for a single response. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-privacy

D is incorrect: Copilot uses a shared AI model across tenants — it is not a separately trained model per tenant. Tenant isolation is achieved through logical access controls and Microsoft Entra authorization at the data layer, not through separate model instances. Reference: https://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-architecture