Meta Tests Facial Recognition to Block Celebrity Scams, Help Users

Meta, the parent company of Facebook and Instagram, is testing new facial recognition technology in an effort to curb the rise of "celeb-bait scams" and help users recover compromised accounts more quickly. This move comes as public figures, like Tom Hanks, have faced increasing challenges with fraudulent AI-generated videos and impersonation attempts.

Targeting "Celeb-Bait" Scams

In a blog post published Monday, Meta outlined its plan to use facial recognition to detect ads that exploit the likeness of celebrities. The system uses facial recognition technology to compare the images in suspected ads with public figures' Facebook and Instagram profile pictures. If the system identifies a match and determines the ad to be a scam, it will block the content and immediately delete any facial data used in the process, whether or not our system finds a match. Meta emphasized that this data is used only for the one-time comparison and is not stored or used for other purposes.

Early tests with a small group of public figures have shown success in identifying and removing these scams. Meta plans to expand the protection, notifying more celebrities that they are automatically enrolled in the system, though they can opt out at any time via the Accounts Center.

Supporting Account Recovery with Facial Recognition

Meta's efforts aren't limited to celebrities. The company is also testing video selfies as a tool for anyone trying to regain access to a compromised account. Users will upload a video selfie, which is then compared to their profile pictures using facial recognition technology. This method, similar to identity verification used in smartphones and apps, offers a quicker and more secure way for users to regain access to their accounts. Meta ensures that all facial recognition data is immediately deleted after the verification process.

This new recovery method aims to provide a safer and faster alternative to traditional document-based verification, which can be more easily exploited by hackers. Meta also reiterated that any video selfies uploaded will be securely encrypted and will not be visible to anyone on the platform.

Strengthening Defenses Against Scammers

Scammers frequently evolve their tactics, and Meta is continually developing new methods to stay ahead. In addition to facial recognition, the company’s automated ad review system already uses machine learning classifiers to analyze ad content and detect scams. Meta is also looking to enhance its tools to detect fake accounts posing as public figures. For instance, scammers might falsely claim that a celebrity is backing a particular investment or request sensitive personal details in return for a free giveaway.

To ensure user privacy and security, Meta has vetted these new measures through a rigorous risk and privacy review process. The company remains committed to exploring innovative approaches to protect both regular users and public figures from online scams while maintaining ongoing discussions with regulators and privacy experts.

Looking Ahead

As Meta continues to test facial recognition technology, it hopes to create a more secure environment on its platforms by improving scam detection and account recovery processes. While challenges remain, this initiative marks a significant step in addressing the growing problem of online impersonation and fraud.