Meta Faces Lawsuit Over AI Training Data and Facial Recognition
Parents allege Meta illegally used their photos for AI training and face recognition without consent.
1Lawsuit Allegations
A group of parents in Illinois and California filed a federal lawsuit against Meta, accusing the company of illegally using their Facebook and Instagram photos to train generative AI models like Emu and Muse Image, and to develop NameTag, an unreleased facial recognition system for smart glasses. The lawsuit claims Meta violated Illinois and California privacy laws by extracting biometric data without consent. The plaintiffs seek damages under Illinois’ Biometric Information Privacy Act, which allows for $5,000 per intentional violation.
2NameTag's Hidden Code
WIRED reported in June that NameTag's code was secretly embedded in Meta’s smart glasses companion app, downloaded over 50 million times. The system was designed to convert faces into biometric signatures and match them against a local database of 'faceprints,' which could receive updates from Meta. While the feature was inactive, researchers confirmed its functionality. Meta later removed the code but denied building a 'universal face database.'
3Biometric Data Sources
The lawsuit suggests Meta derived faceprints from Facebook and Instagram images, citing internal claims that NameTag could recognize people via Meta connections or public Instagram accounts. A Meta patent also describes matching faces against profile photos. The company has not disclosed which images were used for biometric data, leaving plaintiffs to speculate about the scale—potentially millions of users nationwide since September 2021.
4AI Training Controversy
Meta’s generative AI models, like Emu and Muse Image, were trained on vast quantities of Facebook and Instagram images. Muse Image faced backlash this summer for allowing users to generate images based on public Instagram accounts—a feature quickly removed. The lawsuit alleges this training process illegally harvested biometric data from people in the images. Meta defends its transparency but avoids specifics on opt-in policies.
5Meta's Legal History
This isn’t Meta’s first biometric data scandal. In 2020, it paid $650 million to settle an Illinois class action over facial recognition. In 2021, it shut down the system and deleted over a billion faceprints. In 2024, Meta paid Texas $1.4 billion for unlawful biometric collection. The new lawsuit frames NameTag as part of a long pattern of privacy violations, referencing a 2004 Mark Zuckerberg chat mocking users who trusted him with their data.
6Plaintiffs' Stance
The plaintiffs include Francisco Alvarez and his son (Illinois) and Jeremy Wahl and his 10-year-old daughter (California). They seek damages under Illinois’ BIPA ($5,000 per intentional violation, $1,000 per negligent violation) and California laws. The proposed class could cover millions of users whose images were uploaded to Meta platforms or used in AI prompts since 2021. Their attorney argues biometric misuse shouldn’t be a risk of using social media.
7Meta's Defense
Meta calls the lawsuit 'without merit,' emphasizing transparency in AI development. It denies building a central face database and states no final decision has been made about NameTag. CTO Andrew Bosworth initially dismissed WIRED’s report as 'misleading' but later described NameTag on a podcast as a potential feature to recognize people a wearer had met—framing it as exploratory, not consumer-ready.
8Broader Implications
The case highlights growing scrutiny of AI ethics and biometric privacy. If successful, it could set precedents for how companies collect and use personal data for AI training. With Meta’s history of fines and opaque practices, the lawsuit underscores the need for clearer regulations and corporate accountability in the era of generative AI and facial recognition technologies.
Original article: Wired — Meta Faces Lawsuit Over AI Training Data and Facial Recognition