Zoom has partnered with World, the human identity verification company founded by Sam Altman, in a move aimed at curbing the rising threat of AI-generated impersonation in virtual meetings.

The collaboration is designed to help ensure that participants in Zoom meetings are verified humans rather than sophisticated deepfake avatars, as concerns grow over the misuse of artificial intelligence to mimic executives and colleagues.

The risk is no longer theoretical. In early 2024, global engineering firm Arup lost $25 million after an employee in Hong Kong approved a series of wire transfers during what appeared to be a routine video conference with the company’s chief financial officer and several colleagues.

Investigations later revealed that every participant on the call, except the employee, had been generated using AI deepfake technology.

A similar incident targeted a multinational company in Singapore in 2025, underscoring how quickly the technology is being weaponised in corporate fraud schemes.

Financial losses linked to deepfake-enabled fraud surpassed $200 million in the first quarter of last year alone, according to one estimate, while security industry reports put the average loss per corporate incident at more than $500,000.

Although most individuals may never encounter deepfake video-call scams personally, the threat poses a significant risk to companies, particularly those that routinely execute high-value transactions over virtual meetings.

World said that existing safeguards largely focus on scanning video frames for visual clues of AI manipulation, an approach it argues is limited in scope and effectiveness.

Both companies said advances in AI video generation are rapidly eroding the reliability of traditional frame-by-frame deepfake detection systems, making it harder to distinguish between authentic participants and synthetic imposters.

Under the new feature, World deploys its World ID Deep Face technology, which uses a three-layer verification process to confirm that a meeting participant is a real person.

The system cross-checks a cryptographically signed image captured during the user’s initial registration via World’s Orb device, compares it with a real-time facial scan from the user’s device, and matches both against the live video feed visible to other participants in the meeting.