Gartner® names Reality Defender a Market Shaper in deepfake detection, June 2026

Get the report

\

Insight

\

Reality Defender brings real-time deepfake detection to the Microsoft Teams partner ecosystem

Bennett Morrison

VP of Product

An unexpected email asking you to approve a payment would make you stop and verify who sent it. In a live meeting, seeing a familiar face and hearing a familiar voice can make it feel like verification already happened. Meeting controls check who can join the meeting, not who’s actually showing up. That matters because finance teams approve payments, IT teams discuss requests for access, and recruiters decide which candidates to advance. In hiring alone, Gartner predicts that by 2028, one in four candidate profiles worldwide will be fake.

As more critical decisions happen in virtual meetings, teams need a way to spot possible impersonation while there’s still time to verify who’s on the other end. Real-time deepfake detection adds a check inside the meeting itself, alerting the organizer when audio or video shows signs of AI manipulation.

Why Microsoft is bringing detection partners into Teams

At Microsoft Ignite this week, Microsoft named Reality Defender as one of four synthetic media detection partners for Microsoft Teams.

Microsoft calls synthetic media detection one of the hardest problems it’s trying to solve, and it chose to work with specialist detection providers rather than build alone. Partners supply the detection models, and Microsoft supplies the platform, admin controls and in-meeting alerts, so organizations can choose a provider without leaving Teams.

A familiar face on a video call is one of the most convincing things an attacker can fake right now. People trust what they see and hear, and that’s exactly why verification needs to live inside the tools we already use. Getting to do that with Microsoft in Teams is a big deal for us.

Ben Colman, CEO at Reality Defender

How RealMeeting works in Teams

With RealMeeting, the organizer can see detection results privately without leaving the meeting. If RealMeeting flags possible manipulation, the organizer can follow their organization’s policies to pause an approval, confirm a request through a trusted channel, or take additional steps to verify a candidate’s identity.

IT admins can make RealMeeting available across the organization or limit it to groups such as finance or recruiting. Once the app is available, organizers can turn it on in Teams for the meetings that need it. Participants join as usual, with no enrollment or voiceprint required.

RealMeeting analyzes short samples of meeting media for signs of AI manipulation as the conversation unfolds. The organizer sees the detection result and a confidence score in a private Teams side panel. If a scan flags possible manipulation, the organizer can respond in the moment according to their organization’s policies. After the meeting, authorized security teams can review the scan results in the Reality Defender dashboard and investigate any flags. Organizations with strict data requirements can turn on zero retention, which deletes meeting media after analysis.

Security teams can set a consistent standard for handling suspected impersonation across departments, and employees can keep doing sensitive work in the tools and workflows they already know.

Explore RealMeeting on Microsoft Marketplace or contact us to learn more about bringing real-time deepfake detection to your Microsoft Teams meetings.

Frequently Asked Questions

Frequently asked questions

RealMeeting brings deepfake detection directly into Microsoft Teams meetings. Organizers get results in seconds in a private panel and can rescan a participant or end the meeting while it’s still live.

No. Detection runs on the media itself. Participants join as usual, with no enrollment, voiceprint, or PII required.

Admins deploy it organization-wide or to specific groups, and organizers turn it on for the meetings that matter. It rolls out by policy, with no new infrastructure and no change to how people meet.

Full results, including where manipulation was detected, flow to the Reality Defender dashboard for security investigation and review.

The result indicates whether the scan found signs of AI manipulation. An organizer can use that signal to decide whether to verify the participant’s identity or confirm a request through their organization’s existing procedures.