Security/ deepfakes · zoom · biometrics · world

World Adds a Human Badge to Zoom Calls, Backed by Iris Scans

World's new Zoom integration swaps deepfake detection for iris-based proof of humanity, though pricing and rollout details are not public.

Zoom enterprise customers can now put a "Human Badge" next to a colleague's video feed, courtesy of World, the biometric identity company co-founded by Sam Altman. The badge is meant to show that a real person is on the call, not an AI-generated stand-in.

World ID for Zoom is now generally available to enterprise customers. Setup is not a checkbox. Each worker has to visit one of World's Orbs, its iris-scanning devices, which create a cryptographically secure biometric reference (a stored mathematical record of the eye). Companies can choose to have an Orb installed onsite. During calls, a selfie taken in the mobile app is checked against that reference, and verified participants get the badge.

The pitch rests on a shift in strategy. World said deepfake detection can lead to "an arms race between deepfake detection and creation," so it is verifying humans instead of hunting fakes. Microsoft has taken the other route in Teams, opening the platform to third-party deepfake detection tools.

The threat is measurable. Gartner research found that 36% of senior cybersecurity leaders reported at least one deepfake incident during a video call in the past 12 months.

The approach has a logic: proving something is real is a narrower task than proving everything else is fake. It also has a conflict worth noting. Altman runs OpenAI, whose tools are increasingly good at mimicking people, and World sells the remedy.

Several questions are open. World has published no pricing that we could find, and whether an onsite Orb is an expensive option is unconfirmed. We also have no statement from World or Zoom on cost, regional availability, or exactly how the badge behaves on a call. And the plan asks employees to hand over an iris scan to their employer's chosen vendor. Many companies will weigh that request against a 36% incident rate before deciding which of the two they fear more.

TR

The Revision

Written by an AI system from the public sources credited above. How we write →