Meta is facing a proposed class action alleging that the company improperly used Facebook and Instagram photos to train artificial intelligence image models and potentially build biometric data for an unreleased facial-recognition feature designed for its smart glasses, according to WIRED.
The lawsuit was filed in federal court in Chicago by parents and children from Illinois and California.
The plaintiffs allege Meta extracted biometric information from photos without obtaining the notice and consent required by privacy laws in those states. The allegations have not been proven in court.
A major focus of the complaint is NameTag, an unreleased facial-recognition technology discovered in code embedded in Meta’s AI companion application for its smart glasses.
The app had been downloaded more than 50 million times. WIRED previously reported that the system was designed to convert faces captured by the glasses into biometric signatures and compare them against stored faceprints.
The complaint alleges those faceprints may have been derived from Facebook and Instagram images, although Meta has not publicly disclosed what images, if any, were used to create the biometric data.
The lawsuit also challenges Meta’s training of its Emu and Muse Image generative AI systems, alleging that biometric information was extracted from people depicted in training images.
Meta disputes the allegations and says the lawsuit mischaracterizes its work. The company has said NameTag has not been released to consumers and that it is not creating a universal facial-recognition database.
The proposed nationwide class could potentially include millions of people whose images appeared on Facebook or Instagram or were submitted to Meta’s generative AI products.
The complaint seeks damages and injunctive relief, including claims under Illinois’ Biometric Information Privacy Act.
