Loading market data...

World Expands Proof-of-Human ID to Banks, Citing $40 Billion AI Fraud Risk

World Expands Proof-of-Human ID to Banks, Citing $40 Billion AI Fraud Risk

World, the identity project co-founded by Sam Altman, is bringing its proof-of-human verification technology to banks. The company says the bank-focused offering is a response to what it describes as a $40 billion AI fraud risk, and it has signed Bain Capital, VanEck and Pantera Capital as design partners.

The move puts World's iris-scanning identity system in front of financial institutions — a different customer base than the individual users it has targeted since launch.

What the bank offering does

World's core product is a digital ID that verifies a person is human, not a bot. For banks, the company is pitching a version that lets customers delegate proofs to AI agents. That means an AI assistant could confirm a user's identity to a bank without the user having to step in each time.

The design partners — Bain Capital, VanEck and Pantera Capital — are working with World on how the offering will be structured. Their involvement suggests the product is aimed at firms that need to verify customers or counterparties in automated workflows, though World hasn't detailed exactly how each partner will use the technology.

Why World is targeting banks now

World frames the bank push around AI-driven fraud. The company cites a $40 billion figure tied to fraud risk from AI, though it hasn't broken down how that number was calculated. The argument is straightforward: as AI agents become more common, banks need a way to tell a real person from a synthetic one.

That's the same problem World has been trying to solve since it started handing out its orb devices for iris scans. The difference is the audience. Banks have stricter compliance requirements than individual users, and they're more likely to pay for identity infrastructure that can be audited.

The delegation feature

The most notable part of the bank offering is the ability to delegate proofs to AI agents. In practice, a user could authorize an AI agent to act on their behalf — say, to open an account or approve a transaction — while the bank still gets a verified confirmation that a real human is behind the agent.

That's a tricky technical problem. It requires a way to link an AI agent's actions back to a verified human without exposing more personal data than necessary. World hasn't said how it plans to handle that link, or whether the delegation feature will be available at launch or rolled out later.

World hasn't announced a launch date for the bank offering or named any banks beyond the three design partners. The company also hasn't said which markets it will target first or how regulators will treat the delegation feature.

For now, the design partnerships are the clearest signal of how World plans to move beyond individual users. Bain Capital, VanEck and Pantera Capital are all financial firms with their own compliance needs, and their involvement will test whether World's proof-of-human system can meet the standards banks actually require.