World, the biometric identity network backed by Sam Altman and developed by Tools for Humanity, has launched a set of verification products for banks and other financial companies. The tools are intended to help firms determine whether a real person is behind a wire transfer, a video conversation or an instruction issued by an artificial-intelligence agent.
The launch targets fraud involving deepfakes, synthetic identities and software that can act without a person directly carrying out each step. Deloitte has projected that AI-enabled fraud losses in the United States could reach $40 billion by 2027, up from $12.3 billion in 2023. The projection includes losses linked to manipulated media, fabricated identities and autonomous AI agents that can initiate financial transactions.
Three verification products
World said its suite is built on cryptographic technology, which uses mathematical proofs to verify information. Its initial design partners include Bain Capital, VanEck Funds, Pantera Capital, OpenExchange and Outtake. The products are aimed at different points where financial institutions may need to confirm a person’s identity or authority.
| Product | Intended use |
|---|---|
| World ID for Zoom | Displays a “Human Badge” for people who have verified their identity through World during video calls. |
| World ID for Agents | Allows an AI agent to provide cryptographic proof that it is acting for a named, verified person. |
| World ID app | Offers proof that a person is human, present at the time of verification or meets a specified attribute. |
The Zoom tool is designed for high-value or sensitive conversations. World describes it as a way to help counterparties determine whether they are speaking with a person rather than an AI-generated avatar. The company also connects the product to wire fraud and impersonation attacks in which criminals pose as executives or other trusted officials.
The agent product addresses a newer problem: an automated system may be authorized to carry out a task, or it may be controlled by an attacker. World’s system is intended to let an agent show that it is operating for a verified human without disclosing that person’s underlying identity. That distinction could matter when an automated system has access to financial accounts, databases or transaction tools.
What the app verifies
The World ID app is available on iOS and Android in more than 150 countries. It supports three types of proof:
- Proof of Human: a one-time scan at a physical Orb device.
- Proof of Presence: a real-time liveness check performed on a mobile device.
- Proof of Attributes: disclosure of a specific detail from a government-issued credential, such as nationality or whether a person meets an age threshold, without revealing the full credential.
A physical Orb is now available at the New York Stock Exchange, according to The Fintech Times. The announcement does not describe a national requirement for banks, payment companies or consumers to use World’s system. It presents the tools as options for financial institutions seeking additional ways to authenticate people and automated instructions.
The product also raises questions about how identity checks and biometric systems fit with privacy protections. World’s stated approach is to disclose only a requested attribute in some cases rather than the entire government credential. Financial institutions would still need to determine how such proofs fit into their own fraud controls, authentication procedures and obligations for handling consumer information.
