Zoom adds deepfake detection as video call scams get harder to spot

A video call that looks and sounds like a real colleague is no longer proof that the person on screen is real. According to KTLA, Zoom used its Zoomtopia 2026 event to unveil deepfake detection built into its video meetings, a response to how convincing AI-generated faces and voices have become on calls.

The feed summary available from KTLA is short on detail about how the detection works. It frames the update as part of a wider push, features meant to help users look better, sound clearer and get more done, alongside the security addition. But the decision to build deepfake detection into the product at all says something on its own: Zoom is treating a manipulated face on a video call as a mainstream risk, not an edge case.

Why a video call became a weak point

For years, a video call was treated as stronger proof of identity than a phone call or an email. Seeing a face and hearing a voice felt harder to fake than text on a screen. That assumption is what the scam depends on. Generative tools have made it possible to mimic both the look and the sound of a real person in real time, which turns the video call itself into the point of deception rather than a safeguard against it.

That shift matters most in situations where a video call is used to authorise something: a payment, a password reset, an urgent request from someone who appears to be a manager or a family member. If the face and voice on the call can be synthetic, the call stops being a reliable check and becomes another channel that needs its own verification.

What the update does not resolve

The KTLA summary does not say how Zoom's detection works, how accurate it is, or whether it flags a suspected deepfake to the user during the call or afterwards. It is not known, from what KTLA has published so far, whether the feature works on both audio and video manipulation or on video alone.

What is already clear is that the burden of spotting a fake has started to shift from the person on the call to the platform carrying it. That is a meaningful change, but it does not remove the need for a separate check on any call that asks for money, credentials or urgent action, regardless of how convincing the face on screen looks.