Distinguishing a real person from a clever AI bot has become one of the most urgent problems of the AI age—and the stakes keep rising. In recent years, the inability to tell humans and machines apart has fueled billions of dollars in fraud, as AI agents trick companies into providing goods and services. In the most extreme cases, criminals have deployed deep fakes of corporate executives to dupe employees into sending fraudulent wire transfers—a threat that has grown sharper as generative AI tools make convincing synthetic voices and video simple and cheap to produce.

A remedy moved closer to mainstream adoption on Sunday, when World—the company behind an orb device that uses iris scans to establish a “proof of humanity”—announced that video conferencing service Zoom and several big financial companies are adopting its technology to keep out malicious bots. The verification tool is part of a broader suite of services called World ID. For Zoom, a service that became the default venue for board meetings, client pitches, and investor calls during the pandemic-era shift to remote work, the integration marks a step toward treating human verification as routine business infrastructure rather than a novelty.

“World ID for Zoom is designed to protect sensitive, high-stakes conversations by proving that a real and unique human is present. The integration uses cryptography to confirm that the person behind the screen is a verified human,” World said in a statement. The company, which was co-founded by OpenAI CEO Sam Altman, published details of the integration on its official World ID for Zoom page.

Under the Zoom integration, participants who have cleared World ID's security hurdles appear on screen with a small blue badge signaling that they are human and that they are who they claim to be. A screenshot supplied by World shows how the badge appears during a call. Companies, meanwhile, can decide when such proof is required and reserve this level of scrutiny for sensitive conversations.

An Orb, an Iris, and a Tamper-Proof Record

The technology underpinning World ID is controversial, but its advocates say it is effective from both a security and a privacy perspective. Verification is tied to an orb device that captures an image of a person's irises—a biometric identifier distinct to every individual—and creates a permanent, tamper-proof record of their identity.

World's orbs, which are roughly the size of a football, may strike some people as a form of sci-fi dystopia. The devices, however, rely on a type of advanced cryptography that does not create a central database of individuals; instead, they encrypt the identity of those they scan in a way that cannot be read by the company. When an individual needs to verify their identity on Zoom or another platform, they use their phone camera and an app that cross-checks against the encrypted orb record. The design speaks to a core anxiety around biometrics: unlike a password, an iris cannot be changed if it is ever compromised.

According to Tiago Sada, World's chief product officer, consumer reaction to the orb has been similar to the rollout of earlier biometric security measures that took time to gain trust. Sada compared the orb's use of iris scans to when Apple first introduced Touch ID and then facial verification for the iPhone—features that offer a convenient form of security but that many people greeted with suspicion when they first launched.

Sada also noted that the protocol underlying World ID is built on open source code that anyone, including competitors, can use. That openness may ease the concerns of privacy advocates who are uneasy not just with the technology, but with the fact that it is tied to Altman.

“People understand you don't have to give Sam or any of us your photo … Everything is open source and auditable,” Sada said. He added that World ID uses blockchain technology to facilitate decentralization and encryption, observing, “I think 90% of the crypto space is BS, but 10% is profound.”

A Tipping Point for AI Agents

Zoom is not the only big company turning to World's verification tool to stop deep fakes. Also on Sunday, World announced that a host of major financial firms—including Bain Capital and VanEck Funds—are adding the tool to their workplaces. A company spokesperson explained that the financial firms will keep orb devices in their offices so that employees who have not previously used one can scan their irises. For industries where a single impersonated executive can set a fraudulent wire transfer in motion, workplace-level verification offers an additional line of defense.

The spokesperson emphasized that not every verification process requires an eye scan. World ID offers three tiers of verification through its app. The lowest, “Proof of Attributes,” verifies a person's identity based on scanned government documents. The second level, “Proof of Presence,” requires a selfie. The highest security setting, “Proof of Human,” requires a person to verify who they are using the World app and a selfie, which is checked against the biometric record registered with the orb device.

The tiers reflect a broader shift: in an online world increasingly dominated by bots, people may need to verify that they are human before engaging in a widening range of activities, from video calls to dating apps to shopping.

Online shopping is a case in point. Some large companies, notably Amazon and JPMorgan, have already begun to block bots altogether. While such “block all bots” tactics may prove an effective way to ensure that only humans can carry out transactions on their websites, Sada argued that the approach will limit those companies' ability to compete in the fast-growing world of agentic commerce, in which AI agents browse, compare, and transact on their owners' behalf.

To that end, World is also supplying tools that allow users to assign their verified status to an agent under what Sada compares to a power-of-attorney type of arrangement. He stressed that World users will only be able to assign credentials to a single, trusted agent that merchants can treat as their alter ego.

World, formerly known as WorldCoin, is not the only company working on how to credential trusted agents. In August, online security firm Cloudflare ann an agent ID tool of its own, along with a wallet designed for those agents to use. According to Cloudflare Chief Strategy Officer Stephanie Cohen, the new products coincide with an evolution in the digital world in which bots—historically regarded as a menace—are now also treated as a major business opportunity.

It remains to be seen how quickly tools like “Proof of Human” gain real-world traction—and whether other workplace software makers follow Zoom's lead. But as bots continue to proliferate at an ever-growing rate, they appear likely in time to become a permanent part of the digital infrastructure.

This story was originally featured on Fortune.