Meta is facing a fresh privacy lawsuit over its AI-powered Meta Glasses, coming not long after the company agreed to pay up to $18 billion over the next decade to settle child-safety claims.
The case originated in federal court in California in March. An amended complaint filed August 31 expanded it to include bystanders who never bought or wore the glasses but were recorded by them without their knowledge.
The amended complaint alleges Meta marketed the glasses as 'designed for privacy, controlled by you,' while recordings were sent to third-party contractors who could view and label the material for AI development. 'Even if a bystander notices the Glasses and agrees to be recorded, he or she cannot have consented to the mass amalgamation of their data when the Glasses are in AI mode,' the complaint states, adding that visual recordings were 'collected, stored, exploited, and visually inspected by persons overseas' without disclosure or consent from bystanders.
Ryan Clarkson, founder of Clarkson Law Firm and co-lead counsel in the litigation, said corporations are racing to feed more raw material into their AI systems. 'The way that they believe is the best way to improve their AI is to find more signal, to find more content, more recordings that they can feed into this machine, feeding the beast,' he said. He argued the hardware and software together build a 'surveillance economy' driven by 'power' and 'money.'
The glasses combine cameras and microphones with the wearer's field of view, letting users ask questions about what they see, translate text and capture hands-free photo and video. Meta's product website says a capture LED lights up during recording and that users can manage or delete photos, videos and voice interactions.
According to the complaint, when users activate AI features, imagery and audio can be transmitted to Meta's servers for analysis and, plaintiffs allege, used to train Meta's AI models — the practice bystanders have dubbed 'perv glasses.'
The allegations trace back to a February 2026 investigation by Swedish newspapers Svenska Dagbladet and Göteborgs-Posten, which drew on interviews with workers at Sama, a Kenya-based outsourcing firm that performed data-annotation work for Meta. According to those reports, workers described reviewing footage that included people changing clothes, using bathrooms and engaging in intimate activity, along with visible financial and personal information. The original complaint cited the reports in alleging some Sama workers saw identifiable faces despite Meta's claimed anonymization measures.
A Meta spokesperson told Fortune, 'We disagree with these allegations and will fight them,' adding that reviewing AI data to improve products 'works the same way as many other companies,' and that the company filters data to remove identifying information.
The numbers come first: an $18 billion settlement bill, a lawsuit spanning two continents of outsourced labor, and a company insisting its practices mirror the rest of the industry. That defense is itself the tell — it concedes that mass data harvesting from unwitting bystanders has become an industry norm rather than an aberration.
Property rights extend to a person's own image and voice, not just their wallet. When a corporation's terms of service quietly override that principle for millions of non-customers who never signed anything, it is not innovation — it is a breach of the basic contract between a company and the public it claims to serve. The market, not just the courts, will ultimately decide whether Meta's users and shareholders are willing to pay that price.
