The company, which three years ago shut down its “face recognition” system, says it will integrate facial recognition tools into those two products to fight spam and help restore access to compromised accounts in a timelier fashion.

“Scammers are relentless and continuously evolve their tactics to try to evade detection, so we’re building on our existing defenses by testing new ways to protect people and make it harder for scammers to deceive others,” Meta wrote in a blog post.

The goal is twofold, the company says. First, scammers often use images of celebrities to fool people into engaging with advertisements in what’s known as a “celeb-bait” scheme. The tool will compare faces in suspected ads with the public figure’s Facebook and Instagram pages. If there’s a match, and the ad is determined to be a scam, it will be blocked.

Meta says early testing has been promising.

What’s likely of more interest to users, though, is the account recovery potential. Facebook and Instagram are testing the process of using video selfies to verify a user’s identity when an account is reported as compromised. Facial recognition technology will compare that selfie to the profile pictures on the account to determine the legitimacy of the claim.

Those video selfies will never be uploaded to profiles or visible to friends, Meta says. And any facial data generated after the comparison will be deleted.

Facebook abandoned its facial recognition program, which suggested name tags for people in pictures, in 2021, deleting over a billion facial recognition templates. That came after watchdogs warned about privacy implications and after regulators in the European Union began to crack down on the practice.

At the time, Meta wrote that the technology could still be useful for identity verification and anti-fraud measures, and said it would “continue working on these technologies and engaging outside experts.”