A Button for Your Doubt: Deepfake Defense Arrives in the Meeting

For two years, the video meeting has been an open door for deepfake fraud — the channel that took $25 million from one firm and rattled boardrooms everywhere. Now the defenders are finally moving into the room. The most telling part of the biggest platform's response isn't a clever detector. It's a button that asks a human to say: something here feels wrong.

What Happened

In early August 2026, Microsoft confirmed it is extending its "Report a concern" security feature into Teams meetings and group calls, with the rollout completing around October. During a live meeting or in the meeting chat, a participant can now flag suspicious activity — phishing, impersonation, scams, social engineering, or a suspected deepfake — directly from inside the call. The report flows to the Microsoft Defender portal and the Teams admin centre, where an organisation's security team can investigate and act. Microsoft frames it plainly: AI voice cloning and synthetic video have made it easier to impersonate people and harder to detect them quickly, so it is giving meeting participants an easy way to raise the alarm.

One detail matters more than the feature itself. It is not an automatic deepfake detector. It does not scan faces or voices and decide who is fake; it lets a human who suspects something send a report for later review. That places Microsoft's move within a wider, two-track shift in defending the video call. On one track, platforms are building native reporting and controls into the tools. On the other, specialist vendors are adding real-time detection: Pindrop's Pulse for Meetings, named one of TIME's Best Inventions of 2025, now runs across Zoom, Webex and Microsoft Teams to verify that participants are real humans, and firms like DeepTrust market similar detection across Google Meet, Zoom, Teams and others. After years of the meeting being undefended, the room is finally being fortified — from two directions at once.

Why It Matters

The video call became a favoured deepfake attack surface precisely because it felt safe: seeing a face and hearing a voice is how we have always confirmed who we are dealing with. That instinct is what real-time deepfakes exploit. So defenders arriving in the meeting is genuinely good news, and overdue. But the shape of Microsoft's contribution — a report button rather than a detector — quietly tells the truth about where things stand: even the largest platform is not promising to catch the fake for you. It is arming your suspicion, not replacing it. That is honest, and it is also a warning about how much still rests on a human noticing.

advertising