Zoom, Bain Capital Adopt Iris-Scan Badges to Block AI Bots

Zoom and major financial firms including Bain Capital and VanEck Funds are deploying iris-scan verification to separate human users from artificial intelligence agents on video calls and in corporate workflows. The rollout, announced Sunday by the biometric identity company World, introduces a visible blue badge on Zoom screens confirming that a participant has passed cryptographic identity checks.

The integration targets a specific vulnerability: deepfake executives and automated bots tricking employees into authorizing wire transfers or accessing sensitive meetings. World ID for Zoom uses cryptography to confirm the person behind the screen is a verified human, according to a statement from World, which was co-founded by OpenAI CEO Sam Altman.

Corporate clients will keep physical orb devices—roughly the size of a football—in their offices to scan employee irises, a World spokesperson said. Those who have already registered can verify themselves using a phone camera and an app that cross-checks against the encrypted record without exposing raw biometric data to the company.

World offers three security tiers rather than forcing every interaction through an eye scan. The lowest tier, Proof of Attributes, relies on scanned government documents. A second level, Proof of Presence, requires a selfie. The highest setting, Proof of Human, demands a selfie checked against the biometric record registered with the orb device.

The architecture attempts to sidestep the central database problem that doomed earlier biometric initiatives. Instead of storing images, the system encrypts identity data using blockchain technology to facilitate decentralization. “I think 90% of the crypto space is BS, but 10% is profound,” World Chief Product Officer Tiago Sada said.

Sada compared the public reaction to the orbs to Apple’s initial rollout of Touch ID and facial verification on the iPhone—technologies that faced suspicion before becoming standard infrastructure. He also noted the underlying protocol is built on open source code auditable by competitors and privacy researchers.

The push extends beyond keeping bots out. Amazon and JPMorgan have already moved to block automated traffic entirely from certain transactions. But blanket bans threaten to choke off agentic commerce, where consumers delegate purchasing decisions to trusted software. To address this, World is developing tools that allow users to assign their verified status to a single AI agent under what Sada described as a power-of-attorney arrangement, giving merchants a way to treat the bot as the user’s alter-ego.

Competitors are racing toward the same gap. In August, online security firm Cloudflare launched its own agent ID tool alongside a digital wallet designed specifically for AI agents. Cloudflare Chief Strategy Officer Stephanie Cohen said the products reflect a shift where bots are treated as a business opportunity rather than solely a threat.