Meta says it is testing facial recognition technology to combat “celeb-bait” scams on Facebook and Instagram, an initiative first outlined on October 21, 2024. The system checks images—particularly ads—for misuse of public-figure likenesses by comparing them against the figures’ official profile photos, aiming to identify and remove fraudulent content more quickly.
This targeted approach marks a limited return to facial recognition for Meta, which shut down its broader Face Recognition system in November 2021 amid privacy concerns and regulatory pressure. The current deployment focuses exclusively on protecting public figures and high-profile individuals from scams, and Meta says it will not be used for general user surveillance or identification.
“This process is done in real time and is faster and much more accurate than manual human reviews,” said David Agranovich, Meta’s director of global threat disruption. According to the company, any facial data generated for these checks is deleted immediately after the match test.
Fraud Prevention Focus
The rollout responds to rising concerns over sophisticated fraud schemes that use AI and deepfake technology to impersonate celebrities in order to promote fake investment opportunities or counterfeit products. Similar tactics have been documented in the financial sector, where manipulated celebrity images are increasingly used to build false credibility.
Meta is piloting the technology in select markets, with expansion plans based on effectiveness and regulatory input. Regulators in South Korea have approved a trial, and in Australia, Meta has introduced additional ad verification requirements for financial services advertisers. The company is also working with financial institutions and other partners to strengthen scam detection capabilities.
Privacy and Scope
Meta’s policy documentation emphasizes that the feature is limited to public-figure protection in scam contexts. While privacy advocates remain cautious about any return to facial recognition, Meta frames this as a narrowly applied security measure intended to protect both public figures and platform users from evolving fraud threats.
Sources: MSN, Meta, The Guardian
–
By Ji-seo Kim
