Zoom introduces World ID verification to ensure that meeting participants are real humans and not deepfakes.

      Summary: Zoom has teamed up with World, Sam Altman’s biometric identity firm, to enable meeting participants to confirm their humanity through World’s Deep Face technology, which matches iris-scanned biometric profiles with live video to show a “Verified Human” badge. This feature addresses the significant issue of deepfake fraud, which caused companies to lose over $200 million in the first quarter of 2025, including a $25 million loss at engineering company Arup. However, World’s iris-scanning Orb system is currently under regulatory scrutiny in several countries, including Spain, Germany, and the Philippines.

      The collaboration allows meeting participants to prove they are actual humans rather than AI-generated deepfakes. The system uses World’s Deep Face technology to cross-check a participant’s live video against their iris-scanned biometric profile, displaying a “Verified Human” badge if the match is successful. Meeting hosts can activate a Deep Face waiting room requiring verification before participants can join, and attendees can request verification from others during a call.

      This feature addresses a real threat that has evolved from a theoretical concern to a costly reality. In early 2024, engineering firm Arup experienced a $25 million loss when an employee in Hong Kong authorized wire transfers during a video call, unaware that all other participants were AI-generated deepfakes of his colleagues, including the CFO. A similar incident occurred with a multinational company in Singapore in 2025. Overall, deepfake-related fraud caused more than $200 million in losses across the industry in the first quarter of 2025, with the average loss per corporate incident exceeding $500,000.

      Verification mechanics

      World’s Deep Face employs a three-part method for verification. It cross-references a signed image captured at the user’s initial registration through World’s Orb device—a biometric scanner that photographs iris patterns—with a live face scan from the user's phone or computer and a video frame visible to other participants. Verification is successful only when all three inputs match. This procedure is conducted locally on the participant’s device, and World asserts that no personal data is transmitted from the phone.

      This approach differs significantly from existing deepfake detection tools on Zoom’s marketplace. Solutions from companies like Pindrop, Reality Defender, and Resemble AI analyze video frames for signs of AI manipulation and flag synthetic media in real-time. Both Zoom and World noted that as video generation models become more advanced, those frame-by-frame detection methods grow increasingly unreliable. Deep Face entirely circumvents this issue by verifying a person's identity against a biometric record instead of trying to ascertain if the video feed is software-generated.

      However, the trade-off is that Deep Face necessitates participants to possess a World ID, meaning they must have visited a physical Orb device for iris scanning. Currently, there are about 18 million verified users across 160 countries and approximately 1,500 active Orbs, which is a small percentage of Zoom’s user base, thus limiting the feature’s immediate applicability. For most meetings, existing frame-analysis tools are likely to remain the more practical choice. Deep Face is primarily intended for high-stakes calls where identity verification justifies the inconvenience of requiring biometric pre-registration.

      The business rationale

      Zoom spokesperson Travis Isaman described the integration as part of the company’s “open ecosystem approach,” allowing customers to establish trust in their workflows based on their specific needs. This positioning is intentional; Zoom is not positioning World ID as the default identity solution but rather as one of several options available in a marketplace that includes various deepfake detection and identity verification technologies.

      For Zoom, the partnership serves as a defensive strategy. The company reported a revenue of $4.67 billion in fiscal 2025, growing modestly by 3%, and it faces the challenge of maintaining its status as the go-to platform for business communication as competitors introduce AI features across their services. In response, Zoom has rolled out AI avatars, an AI-powered office suite, and cross-application AI notetakers. Incorporating human verification aims to address another aspect: positioning Zoom as the trusted platform for sensitive discussions. In a market where a single deepfake call could cost $25 million, establishing trust holds significant commercial value.

      For World, the Zoom integration marks a success in distribution. Formerly known as Worldcoin, the company has faced challenges in moving beyond its initial crypto-adjacent audience. Partnerships with Visa, Tinder, Razer, and Coinbase have broadened the use cases for World ID, but none of those collaborations generate the immediate demand that a corporate security application would. If a company’s treasury department mandates World ID verification for any video call involving wire transfer approvals, this fosters institutional adoption in a way that consumer partnerships do not.

      The privacy concerns

      World's Orb-based identity verification system has come under prolonged regulatory examination. Spain’s data protection authority issued a formal warning in February 2026 for GDPR violations and insufficient data protection evaluations. In December 2024, Germany’s Bavarian data protector ordered the deletion of iris data. The