Facebook parent company Meta Platforms Inc. will start using facial recognition technology to crack down on scams that use pictures of celebrities to look more legitimate, a strategy referred to as “celeb-bait ads.”
Scammers use images of famous people to entice users into clicking on ads that lead them to shady websites, which are designed to steal their personal information or request money. Meta will start using facial recognition technology to weed out these ads by comparing the images in the post with the images from a celebrity’s Facebook or Instagram account.
[time-brightcove not-tgx=”true”]
“If we confirm a match and that the ad is a scam, we’ll block it,” Meta wrote in a blog post. Meta did not disclose how common this type of scam is across its services.
With nearly 3.3 billion daily active users across all of its apps, Meta relies on artificial intelligence to enforce many of its content rules and guidelines. That has enabled Meta to better handle the deluge of daily reports about spam and other content that breaks the rules. It has also led to problems in the past when legitimate accounts have been unintentionally suspended or blocked due to automated errors.
Read More: The Face Is the Final Frontier of Privacy
Meta says it will also start using facial recognition technology to better assist users who get locked out of their accounts. As part of a new test, some users can submit a video selfie when they’ve been locked out of their accounts. Meta will then compare the video to the photos on the account to see if there is a match.
Meta has previously asked locked-out users to submit other forms of identity verification, like an ID card or official certificate, but says that the video selfie option would only take a minute to complete. Meta will “immediately delete any facial data generated after this comparison regardless of whether there’s a match or not,” the company wrote in a blog.
The social networking giant has a complicated history with facial recognition technology. It previously used facial recognition to identify users in uploaded photos as a way to encourage people to tag their friends and increase connections. Meta was later sued by multiple U.S. states for profiting off this technology without user consent, and in 2024 was ordered to pay the state of Texas $1.4 billion as part of the claim. Several years earlier, it agreed to pay $650 million in a separate legal suit filed in Illinois.
The company will not run this video selfie test in Illinois or Texas, according to Monika Bickert, Meta’s vice president of content policy.