Building, integrating, or customizing apps and workflows within Microsoft Teams using developer tools and APIs
Based on your description, I understand that you are trying to programmatically extract and track usage metrics for the Microsoft Teams Copilot Facilitator, specifically focusing on metrics like meeting recap generation, transcript creation, and AI-generated meeting notes, using the Microsoft Graph API and the Office 365 Management Activity API.
Currently, your main roadblock is that standard events (like Audit.General, Audit.SharePoint, or LiveNotesUpdate) do not offer a granular or distinct way to explicitly verify if Copilot Facilitator was active inside a given meeting.
Below is a detailed breakdown answering your specific questions, highlighting a promising audit record type discovered for Team Copilot interactions, and outlining technical workarounds you can leverage right now.
- Direct APIs & Dedicated Audit Event Indicators (Current State)
Based on public Microsoft documentation, there is no single out-of-the-box flag or boolean API property (such as FacilitatorUsed = true) that directly states whether Copilot Facilitator was initialized during a specific Teams meeting. However, you can piece this information together by targeting specialized telemetry and backend APIs:
While standard audit categories are broad, Microsoft logs deep-level user interactions with Copilot in Teams under a specific record type that you can filter for in your extraction pipelines:
-RecordType: TeamCopilotInteraction (RecordType ID = 334). This specific type is dedicated to user interactions with the Copilot Facilitator and automated actions executed by the Facilitator engine within Teams.
-Target Operations to Filter:
-
AINotesUpdate -
LiveNotesUpdate -
TeamCopilotMsgInteraction
Key Field Verification: Within the payload JSON, inspect the AppIdentity property. For Facilitator-specific content, this identifier typically starts with or contains Copilot.TeamCopilot.
Note: Microsoft reminds developers that Purview audit logs are primarily architecture-designed for security and compliance auditing rather than strict business usage reporting. Consequently, metrics aggregated purely from audit logs may slightly deviate from telemetry found in Admin Center dashboards.
- Utilizing the Microsoft Graph Meeting AI Insights API
If you need to query the actual outputs generated by the AI for a meeting (which implicitly validates its usage), you should leverage the Meeting AI Insights API. If a meeting has a transcript and Copilot was utilized, you can pull the AI-generated context using these endpoints:
-
GET /copilot/users/{userId}/onlineMeetings/{onlineMeetingId}/aiInsights -
GET /copilot/users/{userId}/onlineMeetings/{onlineMeetingId}/aiInsights/{aiInsightId}
This returns structured blocks containing:
- Meeting Notes (AI-generated summary text)
- Action Items (Assigned tasks captured by the AI)
- Participant Mentions
Please refer the following document: Get AI-generated meeting summaries with Meeting AI Insights API
- Aggregated Reporting Endpoints
For broad operational oversight, adoption metrics, and organization-wide trends, Microsoft exposes metadata through:
- Copilot Usage Reports API: Provides active user counts and application breakdowns (Word, PowerPoint, Teams), but lacks specific granular row-by-row meeting IDs.
- Microsoft 365 Admin Center Reports & Viva Insights: Offers analytical reporting on overall Copilot adoption and meeting hours saved, but does not expose an external raw query mapping back to individual private meeting artifacts.
Please refer the following document: Microsoft 365 Copilot APIs overview
Currently, Microsoft has not issued a public document or timeline for releasing a dedicated feature flag endpoint specifically for "Facilitator Usage Tracking." Instead, engineering focus remains heavily centered on extending the Meeting AI Insights API and the AI Insights Change Notifications API (currently in preview), which allows applications to subscribe to webhooks whenever new AI-generated meeting artifacts are compiled.
I hope this information help.
If the answer is helpful, please click "Accept Answer" and kindly upvote it. If you have extra questions about this answer, please click ""Comment"".
Note: Please follow the steps in our documentation to enable e-mail notifications if you want to receive the related email notification for this thread.