Zoom has partnered with World, the biometric identity company co-founded by Sam Altman, to introduce a verification feature that lets meeting participants prove they are real humans rather than ai-generated deepfakes. The integration relies on World’s Deep Face technology, which matches a participant’s live video feed against an iris-scanned biometric profile captured earlier through the company’s Orb device. When the match succeeds, a “Verified Human” badge appears next to the person’s name. Hosts can require verification in a waiting room before anyone joins, or request it mid-call if doubts arise.
The move comes as deepfake fraud has shifted from a niche concern to a costly reality for businesses. In one widely reported case in early 2024, engineering firm Arup lost $25 million after an employee authorised wire transfers during a video call where the other participants were deepfake versions of his colleagues, including the chief financial officer. Similar incidents followed, pushing industry-wide losses from deepfake-enabled fraud above $200 million in the first quarter of 2025 alone, with the average corporate incident now exceeding $500,000. These figures illustrate how convincingly synthetic video can mimic real people in real-time business settings.
Deep Face takes a different approach from the deepfake detection tools already offered in Zoom’s marketplace, such as those from Pindrop, Reality Defender and Resemble AI. Those systems analyse video frames for signs of artificial manipulation. As video-generation models have improved, however, such detection methods have become less reliable. Deep Face sidesteps the problem by verifying identity against a pre-registered biometric record rather than attempting to spot synthetic pixels. The process runs locally on the user’s device and, according to World, does not transmit personal data.
The catch is that participants must already hold a World ID, which requires visiting one of the company’s physical Orb scanners to have their irises photographed. With roughly 18 million verified users across 160 countries and about 1,500 active Orbs, the network remains small compared with Zoom’s vast user base. For everyday meetings the existing frame-analysis tools will likely stay more practical. The new verification is positioned for high-stakes calls where absolute certainty about identity justifies the added friction of biometric pre-registration.
For Zoom, whose fiscal 2025 revenue reached $4.67 billion with modest 3 percent growth, the partnership forms part of a broader effort to keep the platform relevant amid intensifying competition. The company has rolled out ai avatars, ai notetakers and other productivity features, but trust in sensitive conversations represents a different and increasingly valuable dimension. A single compromised call can now carry multimillion-dollar consequences, making any credible layer of assurance commercially significant.
For World, which rebranded from Worldcoin in 2024 after its early crypto associations, the Zoom integration offers a route into mainstream enterprise use. Previous partnerships with companies such as Visa, Tinder, Razer and Coinbase expanded the contexts in which a World ID might prove useful, yet none matched the institutional pull of a corporate security requirement. If treasury teams begin mandating verification for calls involving financial authorisations, adoption could accelerate beyond individual consumer uptake.
Yet the partnership also revives long-standing questions around World’s iris-scanning system. Regulators in Spain, Germany, the Philippines and several other countries have raised concerns over data protection, consent practices and potential gdpr violations. Orders to delete iris data, cease-and-desist notices and formal warnings have accumulated since late 2024. The eu ai act’s classification of biometric identification as high-risk adds another layer of complexity for organisations operating in Europe. While World emphasises zero-knowledge proofs and local processing, critics point out that the initial collection of biometric data through physical Orb visits carries risks that cryptography alone may not resolve, especially given reports of recruitment focused on lower-income communities.
Enterprises weighing the feature will need to balance the security benefit against these regulatory and reputational considerations. A financial trading desk handling large transactions might accept the trade-off; a European public-sector body almost certainly will not. The calculus will vary sharply by jurisdiction and risk appetite.
In the end, the Zoom-World integration signals how deeply the deepfake problem has embedded itself in business communications. Two years ago the Arup case seemed like an outlier. Today, proving you are genuinely human on a video call has become a practical concern rather than a theoretical one. Biometric verification anchored to iris scans offers one technical answer, but it brings its own frictions around privacy, compliance and accessibility. It is unlikely to become standard for routine meetings, yet its presence on Zoom’s platform reflects a broader shift: in an era of increasingly sophisticated synthetic media, simply seeing someone on screen may no longer be enough to assume they are real.
