Microsoft is introducing a new security feature for Teams that enables the integration of certified third-party tools. These tools are designed to scan audio and video streams during meetings to detect potential deepfake content.
The rollout of this functionality is scheduled to begin in mid-November. It will be available across the desktop version of Teams, the Mac application, and the web interface, providing users with automated warnings.
To implement the detection, developers must utilize the reportSyntheticMedia API, which is currently in beta. Certified bots require specific administrative permissions, specifically Calls.ReportSyntheticMedia.All, to access and analyze meeting data.
Once active, the system assigns a confidence score ranging from 0.0 to 1.0 to participants. It also categorizes the likelihood of synthetic media into three levels: low, medium, and high, allowing for clearer risk assessment.
Microsoft emphasizes that the actual analysis is performed by external providers rather than the platform itself. The company advises administrators to thoroughly review the data processing policies and terms of service of any third-party provider before enabling the feature.
