Meta Expands Anti-Scam Facial Recognition Test to UK, Europe

# meta

Last October, Meta took a tentative step into the realm of facial recognition by testing two new tools internationally. One tool aims to curb scams that use images of celebrities, while the other helps users regain access to compromised Facebook or Instagram accounts. Now, Meta is broadening this experiment to include another significant country. Initially, Meta had excluded the United Kingdom from its facial recognition test, but as of Wednesday, it started rolling out both tools there as well. In countries where these tools are already in use, Meta is expanding the "celeb bait" protection to more users, and the company plans to extend the test across the European Union. Meta received approval to launch in the U.K. after discussions with local regulators, who have been increasingly supportive of AI technologies. "Over the coming weeks, public figures in the U.K. will begin receiving in-app notifications informing them that they can opt into the celeb-bait protection using facial recognition technology," Meta stated. Both this feature and the new "video selfie verification" for all users are optional, according to Meta. While Meta has a history of using user data to enhance its algorithms, when it introduced these facial recognition tests in October 2024, the company emphasized that the features were solely for fighting scam ads and user verification. "We immediately delete any facial data generated from ads after a one-time comparison, whether or not our system finds a match, and we don't use it for any other purpose," explained Monika Bickert, Meta's VP of content policy, in a blog post. This move comes as Meta is heavily investing in AI technologies. Alongside developing large language models and integrating AI into its products, Meta is also reportedly working on a standalone AI app. The company has increased its lobbying efforts around AI and has shared its views on potentially dangerous AI applications—those that could be weaponized—while suggesting that its own AI developments are not risky. Given Meta's past challenges, focusing on tools that address immediate issues within its apps could be a smart way to gain acceptance for new facial recognition features, an area where Meta has faced difficulties. This test aligns with that strategy: Meta has long been criticized for not doing enough to prevent scammers from using celebrities' images to promote fraudulent schemes, like questionable crypto investments. Facial recognition has been a contentious issue for Meta. In 2024, the company settled a lawsuit for $1.4 billion over allegations of improper biometric data collection related to its facial recognition technology. Before that, in 2021, Facebook discontinued its facial recognition tool for photos, which had encountered numerous regulatory and legal hurdles in various regions. However, the company decided to keep the DeepFace model, hinting at its potential use in future technologies, which might be part of the current offerings. *Updated to note that the test will also run in the EU.*

Related article

Comments (48)