Meta to Use Video Selfies for Scam Detection, Faster Account Recovery

Meta is rolling out new security measures to combat scams and improve account recovery for its users, including the use of facial recognition against scam ads that misuse images of public figures, a scheme known as “celeb-bait.”

Scammers often lure unsuspecting users into engaging with fraudulent ads by using images of celebrities or influencers. These “celeb-bait” scams typically redirect users to phishing websites, where they are prompted to share personal information or make financial transactions.

Though legitimate ads often feature well-known public figures, the deceptive nature of celeb-bait ads makes them difficult to distinguish from authentic content.

Meta’s ad review system already uses machine learning to analyze millions of ads daily, scanning for scams or policy violations. However, to enhance detection capabilities, Meta is now testing a facial recognition feature.

If the system suspects that an ad might contain the image of a public figure being exploited in a scam, it will compare the ad’s image to the person’s Facebook or Instagram profile picture. If a match is confirmed and the ad is deemed fraudulent, it will be blocked immediately.

In addition to fighting scams, Meta is focusing on improving account recovery for users who have lost access to their accounts. Common reasons for losing account access include forgetting passwords, losing devices, or falling victim to phishing scams.

When an account is compromised, Meta requires users to verify their identity before restoring access, typically by submitting official identification documents.

To streamline this process, Meta is now testing the use of video selfies for identity verification. Users can upload a short video of themselves, which Meta will then compare to the profile pictures associated with the account in question.

This method, which mirrors other technologies used for phone or app security, provides a faster and more secure way to regain access to compromised accounts.

Meta believes these new measures, especially those involving facial recognition, will provide more robust protection against account recovery abuse compared to traditional document-based methods.