Microsoft Defender XDR
Hunting and detection

SyntheticMediaDetectionEvents table in the advanced hunting schema - Microsoft Defender | Microsoft Learn

In brief

Microsoft added documentation for the SyntheticMediaDetectionEvents table, which provides Microsoft Teams synthetic media detection events, including severity, confidence, maliciousness, source, user, and detection provider details.

What Defender admins need to know

Administrators can use the table to investigate deepfake activity. Results require Defender for Office 365 and an approved detection engine.

Summaries are generated from the documentation change itself.

Documentation change

The comparison below shows only the changed extract. Use the full-page view for complete context.

new file mode 100644

layout: Conceptual title: SyntheticMediaDetectionEvents table in the advanced hunting schema - Microsoft Defender | Microsoft Learn canonicalUrl: https://learn.microsoft.com/en-us/defender-xdr/advanced-hunting-synthetic-media-detection-events breadcrumb_path: /defender-xdr/breadcrumb/toc.json permissioned-type: public feedback_system: Standard feedback_product_url: https://techcommunity.microsoft.com/t5/microsoft-365-defender/bd-p/MicrosoftThreatProtection uhfHeaderId: MSDocsHeader-MicrosoftDefender manager: orspodek description: Learn about the SyntheticMediaDetectionEvents table in the advanced hunting schema, which contains details about synthetic media (deepfake) detection events in Microsoft Teams. ms.service: defender-xdr ms.subservice: adv-hunting ms.author: dhagarwal author: poliveria ms.localizationpriority: medium ms.collection:


SyntheticMediaDetectionEvents table in the advanced hunting schema - Microsoft Defender | Microsoft Learn

The SyntheticMediaDetectionEvents table in the advanced hunting schema contains details about synthetic media (deepfake) detection events in Microsoft Teams in your organization.

This advanced hunting table is populated by records from Microsoft Defender for Office 365. If your organization hasn't deployed the service in Microsoft Defender, queries that use the table don't work or return any results. For more information about how to deploy Defender for Office 365 in the Defender portal, read Deploy supported services.

Note

This table contains data only if you're using an approved detection engine: Microsoft or a third-party engine (GetReal, Pindrop, Reality Defender, or Resemble AI).

Schema

For information on other tables in the advanced hunting schema, see the advanced hunting reference.

Column nameData typeDescription
DetectionIdstringUnique identifier for this detection event
TimestampdatetimeDate and time when the activity was recorded
SeveritystringSeverity of the detection; possible values: High, Medium, or Low
ConfidenceScorerealIndicates whether the synthetic media was likely used with harmful intent (for example, impersonation, fraud, or social engineering), rather than benign use (AI-generated avatars or accessibility tools)
IsMaliciousbooleanIndicates whether the synthetic media was likely used with harmful intent (for example, impersonation, fraud, or social engineering), rather than benign use (AI-generated avatars or accessibility tools)
SourceTypestringThe Microsoft 365 application where the detection event occurred
SourceIdstringUnique identifier of the source (for example, call ID for meetings or calls, network message ID for emails, and file ID for files)
SourceContextUrlstringURL of the source (for example, the meeting URL or file URL)
SourceContextNamestringName of the source (for example, the meeting, file, or email name)
SourceContextTypestringContext of the source (for example, scheduled meeting, ad hoc meeting, or recurring meeting)
DetectedUserIdTypestringType of identifier to interpret DetectedUserId; possible values: PSTN (for phone number), Enterprise, Consumer, or Bot
DetectedUserDisplayNamestringDisplay name of the user flagged as using synthetic media
DetectionProviderNamestringName of the detection bot or application that identified the synthetic media
ModalitystringType of content analyzed; possible values: Audio, Video, Multimodal, or Image
IsRealTimebooleanIndicates whether the detection happened live during the session (true) or after as a post-session analysis (false)
ContentDurationSecrealDuration of the suspicious content segment in seconds
DetectionsdynamicDetailed results from each detection model; each entry includes model name, content type analyzed, detection tasks, time segment, and confidence score
PrimaryModelTasksstringThe primary types of synthetic media detected

Correlate detections with call activities

Join the SourceId column in SyntheticMediaDetectionEvents with the CallId column in the CallActivityEvents table to view activities recorded for the same Microsoft Teams call or meeting.

The following query returns up to 500 matching detection and activity records, starting with the most recent call activities. The inner join returns only records with matching call IDs and preserves multiple detections and activities for the same call.

CallActivityEvents
| join kind=inner (
    SyntheticMediaDetectionEvents
    | where isnotempty(SourceId)
) on $left.CallId == $right.SourceId
| project DetectionTimestamp = Timestamp, DetectionId, Severity,
    DetectedUserDisplayName, CallId, ActivityTimestamp, ActivityType,
    ActivityInitiatorDisplayName
| order by ActivityTimestamp desc
| take 500

\ No newline at end of file