Meta evaluates facial recognition technologies to combat fraudulent celebrity impersonations.
Three years after Meta discontinued face recognition tech on Facebook due to significant privacy concerns and regulatory pressure, the social media company announced on Tuesday that it is re-evaluating the service as part of an initiative to combat “celeb bait” schemes.
Meta announced it will enroll approximately 50,000 prominent people in a trial that entails the automatic comparison of their Facebook profile photographs with images utilized in suspected fraudulent advertisements. If the photos correspond and Meta determines the advertisements to be fraudulent, it will prohibit them.
The company stated that the celebrities will be informed of their membership and may decline participation if they want not to partake.
The company intends to initiate a global trial in December, omitting significant locations without regulatory approval, including Britain, the European Union, South Korea, and the U.S. states of Texas and Illinois.
Monika Bickert, Meta’s Vice President of Content Policy, stated during a briefing with journalists that the business was focusing on public figures whose images it has recognized as being utilized in fraudulent advertisements.
The objective is to implement extensive protection for them. Bickert stated that individuals may choose to opt out if they like, but the intention is to provide this security in a manner that is accessible and convenient for them.
The test demonstrates a corporation attempting to navigate the complexities of employing potentially intrusive technologies to alleviate regulatory apprehensions regarding the increasing prevalence of scams, while also mitigating grievances related to its management of user data, a longstanding issue for social media firms.
In 2021, Meta discontinued their facial recognition system and erased the facial scan data of one billion users, citing “increasing societal concerns.” In August of this year, the firm was mandated to remit $1.4 billion to Texas to resolve a state complaint alleging the unlawful collection of biometric data.
Simultaneously, Meta is confronted with litigation alleging its inadequate measures to prevent celebrity bait frauds, which exploit photos of renowned individuals, frequently created by artificial intelligence, to deceive users into investing in fictitious schemes.
Under the new trial, the company stated it will promptly eliminate any facial data produced from comparisons with questionable advertisements, irrespective of whether a scam was identified.
Bickert stated that the tool underwent Meta’s comprehensive privacy and risk review procedure internally and was also deliberated with regulators, legislators, and privacy experts externally prior to the commencement of tests.
Meta announced its intention to experiment with facial recognition data to enable non-celebrity users of Facebook and Instagram to recover accounts hacked by hackers or closed due to forgotten passwords.