The owners of Facebook, Instagram, and WhatsApp will use facial recognition technology in the European Union, the UK, and South Korea to combat fraudulent ads that use celebrities’ images and to facilitate account recovery with a video to verify identity. Meta warns of a surge in scams using fake celebrities to steal personal data and money, exploiting users’ trust in familiar faces.

However, the feature will be optional and only available for public figures. If activated, Meta’s systems will use this technology to compare the face of a celebrity from the profile image on their official Facebook and Instagram accounts with the one used in the fraudulent ad.

If a match is confirmed and it is determined that the ad is a scam, it will be blocked. The company has explained that this is a real-time process, which is faster and more precise than manual review by a person. Meta has been ramping up its efforts to fight online fraud. The company recently announced that it has removed more than two million accounts linked to the “Pig Butchering” scam in collaboration with international police forces. These scams are operated by criminal organizations that force workers into executing various financial fraud schemes, including impersonation and investment scams. Once victims transfer their money, they are typically unable to recover their funds as the scammers disappear.

Facial recognition to recover accounts

Meta will also use facial recognition to speed up the recovery of a compromised Facebook or Instagram account. In this case, the company explains that they may request a video selfie to verify the identity of the person and confirm that they are a legitimate user. This measure aims to help people in case their account has been hacked, but it will also allow the platforms to verify a person’s identity if the system suspects that someone has accessed an account without consent.

This system asks the user to upload a video of themselves, a video selfie, which the system compares with other profile pictures using facial recognition technology. The process, according to the company, takes around one minute. Biometric security will become a key defence tool for big tech companies. In the next “10 to 15 years the use of facial recognition systems will become a normal part of how we secure ourselves,” said a Harvard Review study in 2022.

How will facial recognition data be stored securely?

Meta assures that this measure has privacy and security guarantees; they only use the video selfies once for identity verification, which is encrypted and securely stored as soon as it is uploaded. The video will never appear published on the user’s profile, and once the identity verification is completed, the company will delete the generated facial data, regardless of whether there was a match or not.

The video selfie will be just one of the options the company offers users to recover their accounts. However, they argue that it is a more robust, precise, and fast system, and more difficult for a malicious actor to abuse because it requires the legitimate owner of the account to perform the recovery. It also requires a real video, making it resistant to videos generated by artificial intelligence.

“Scammers are relentless and continuously evolve their tactics to evade detection. We’re just as determined to stay ahead of them and will keep building and testing new technical defenses to strengthen our detection and enforcement capabilities,” underscores Meta.