Zoom rolls out “Proof of Humanity” to combat deepfake impersonation

We’ve likely all seen by now how convincingly realistic AI content can be, and, as a growing trend, so-called ‘deepfake’ technology is now being used in impersonation scams during video calls.

It’s not a niche operation; as many as half of UK businesses have already been targeted by AI cybersecurity scams, which can have major financial implications. It’s no surprise that businesses are now seriously rethinking their trust in online communications. 

To address the growing risk, Zoom has announced a new “Proof of Humanity” verification feature designed to confirm that meeting participants are who they claim to be, not AI doubles.

The growing fraud risk facing UK businesses

New research from Gallagher recently found that half of UK businesses have been targeted by AI fraudsters impersonating senior executives in the past year, racking up average losses of £750,000 per incident.

The rapid development of generative AI tools means that scammers can pull off more sophisticated operations. Criminals can now convincingly mimic voices and visual likeness to an uncanny degree, partly due to the widespread availability of public information and photos on platforms such as LinkedIn.

After adopting the guise of senior leadership, they’ll then encourage employees to transfer funds or share sensitive company information. 

The survey also reported that over a third of firms have faced serious extortion or impersonation threats, while 51% of directors cite AI-enabled deception as their top concern.

What Zoom’s “Proof of Humanity” does

To counteract the growing risk, Zoom implemented a new feature last week that allows users to prove their humanity.

Developed in partnership with Tools for Humanity, the video conferencing software platform will now allow users to verify their identity in real time during video calls using a technology called “World ID Deep Face”.

The process involves initial enrolment via a specialised device, followed by on-the-spot authentication when joining calls. 

With a somewhat dystopian touch, successfully verified users get to display a “Verified Human” badge during meetings.

The feature is aimed particularly at enterprise and regulated environments, such as financial services or healthcare, where both the risk and potential cost of impersonation fraud are greatest.

Why this matters now

As generative AI tools become cheaper and more readily accessible, the barrier to carrying out sophisticated scams is dropping rapidly. 

As two of the biggest trends in the world of work, remote working and AI converge, this is one of the latest and potentially damaging risks, especially in sectors that have embraced and now rely on video conferencing. 

Industry estimates suggest AI-enabled fraud losses could grow sharply in the coming years, we’re talking billions. So Zoom’s approach to verification signals an important step toward regaining trust in what was once considered a relatively secure mode of communication. 

But concerned business owners can also take additional precautions, such as exploring additional payment verification steps and ensuring staff are aware of these scams to guard against increasingly convincing attempts.