Meta is testing face recognition to battle scams impersonating celebs

Celebrity impersonation became one of the more serious threats on social media to users

Meta, the parent company of Facebook and Instagram, is testing out facial recognition tools to combat a rising problem: fake ads featuring celebrities. These are popularly referred to as “celeb-bait” scams, where scammers use images of popular public figures to dupe users into visiting malicious websites. Meta says this is one way to offer protection for both its platforms and their users against misleading ads, as well as easing account recovery for users.

This new technology that Meta is about to introduce will focus most of its resources on finding phony ads, which feature images of celebrities and are used to lure people into scams. “Celeb-bait” scams have always been a problem on both Facebook and Instagram, where fraudsters try to impersonate public figures to promote phony products or services. Already, Meta uses automated means like machine learning to detect these scams; however, some ads can pass through the system because they closely resemble real promotions.

To combat this, Meta has instituted face recognition technology. If an ad comes in and is flagged suspicious, Meta matches the image on the ad to the celebrity's profile picture on Facebook or Instagram. If that confirms a match, and it is indeed scamming, then Meta blocks it. Early tests showed promising results, and the company promised to expand use of the technology to add more celebrities in the future.

An attempt to quickly block scamming accounts

In addition to preventing scam ads, Meta is working to protect celebrities and influencers from being impersonated on its platforms. Scammers often create fake accounts to pose as public figures, using these profiles to push fraudulent schemes or ask for personal information. By utilizing facial recognition, Meta aims to quickly identify and block these impersonation accounts before they can cause harm.

According to the official press release, Meta said:^[1]

We use machine learning classifiers to review every ad that runs on our platforms for violations of our ad policies, including scams. This automated process includes analysis of the different components of an ad, such as the text, image or video.

Public figures who have been targeted by these scams will be automatically enrolled in Meta's protection program, although they can choose to opt-out if they prefer. This effort not only helps celebrities safeguard their reputation but also protects users from falling victim to fraudsters pretending to be famous people.

Faster account recovery could be one of the reasons for users to accept the technology

Meta's facial recognition tools are not just limited to scam prevention. They are also being tested as a way to help users recover access to their accounts more efficiently. If someone is locked out of their account, they can submit a video selfie, which Meta will compare to the profile picture on their account. This process is faster and more secure than traditional methods like providing identification documents.

The video selfie method is similar to technologies used by other platforms, such as Apple's Face ID.^[2] Meta assures users that the video selfie will be encrypted and securely stored, with any facial data deleted after the comparison is complete. This new account recovery option aims to provide a more user-friendly experience, especially for those who have been hacked or locked out of their accounts.

No testing done on EU users so far

Meta said these new tools have been tested with the “utmost rigor” for security and privacy. Data from facial recognition is used exclusively for doing the job at hand, say identifying a scam or verifying a user, and then instantly deleted. That said, with all these assurances, the rollout of some of these tools has been delayed in certain regions due to strict privacy laws, including the European Union and the United Kingdom.

In the meantime, Meta continues to engage with regulators and policymakers to ensure the tools meet legal requirements. While the new features are gradually being introduced around the world, Meta remains focused on enhancing user protection without compromising privacy. Andrew Devoy, the spokesperson for Meta, said the following:^[3]

We are engaging with the U.K. regulator, policymakers, and other experts while testing moves forward. We’ll continue to seek feedback from experts and make adjustments as the features evolve.