TL;DR
World — the Sam Altman co-founded identity network — unveiled World ID for banking and finance at Money20/20 on October 18, its first industry-specific push. The suite bundles an identity wallet, agent-credential delegation, and deepfake checks for video calls so institutions can tell a verified human (or their agent) apart from anonymous automated traffic. Named design partners include Bain Capital, Pantera Capital, VanEck Funds, OpenExchange, and Outtake. The announcement names no bank customer, and the agent piece is still in beta.
What World announced
On Sunday, October 18, World co-founder Alex Blania took the stage at Money20/20 in Las Vegas to unveil World ID for banking and finance, which the company describes as the first time it has pointed World ID at a single industry. The suite is now available to help banking and financial institutions make transactions, trades, customer requests, and video calls AI-safe, across internal and customer-facing systems.
The bundle has three parts. The World ID app is an identity wallet offering three cryptographic credentials: Proof of Human (a one-time iris verification at an Orb), Proof of Presence (a quick phone liveness check), and Proof of Attributes (confirming a fact like age or nationality from a government credential while revealing only the requested detail). World ID for Agents — AgentKit, currently in beta — lets a person delegate certain proofs to an AI agent, which can then transact on their behalf and prove it is acting at a real human's behest without revealing that person's private information. World ID for Zoom is now generally available after a limited beta announced in April: a verified human gets a visible Human Badge on their Zoom tile, Zoom receives no personal information in the process, and meeting hosts can soon require proof-of-presence checks in the waiting room before admitting participants.
Why banks would care
World is pitching the suite as an answer to AI-driven fraud. It points to the early-2024 case in which an employee at engineering firm Arup wired approximately $25.6 million after joining a video call where every other participant was a deepfake, and to a U.S. Treasury FinCEN warning that criminals are using deepfake photos and video to beat checks on new-account applications. Deloitte's projection — cited in World's announcement — puts AI-enabled fraud losses in the U.S. at $40 billion by 2027, up from an estimated $12.3 billion in 2023.
The agent angle may matter more over time. In its release, World said: “For a merchant or financial institution, that offers a way to distinguish an AI agent connected to a verified human from anonymous automated traffic that some sites are now blocking.” The pitch: as personal agents take on more financial work, delegated proofs become the credential that separates an authorized agent from a bot.
What's still to prove
The named design partners — Bain Capital, OpenExchange, Outtake, Pantera Capital, and VanEck Funds — are financial firms exploring uses, not announced customers. World's announcement names no bank as a customer and gives no pricing or rollout timeline. AgentKit, the piece that lets agents spend with delegated credentials, remains in beta — so the delegation half of the pitch is the least proven.
World says more than 18 million people hold its Proof of Human credential, in line with the count it reported in July. Tools for Humanity, World's main developer, has joined Zoom's independent software vendor exchange, and the company plans to present the technology at Zoomtopia and Cloudflare Connect. What banks do next — pilot, partner, or pass — is the part worth watching.