The legal battle over Meta's camera-equipped glasses just got much bigger — and it no longer involves only the people who bought the device.

In a 230-page consolidated amended complaint, 72 plaintiffs have merged 13 class actions that a US court combined in April. The case, Bartone v. Meta Platforms (3:26-cv-01897-EMC, Northern District of California), was originally about Ray-Ban Meta wearers who said Meta's 'designed for privacy' marketing was misleading. The amendment now adds a second, potentially far larger group: 'bystanders' who never owned the glasses but were filmed or recorded while someone else used them — and whose faces, voices and surroundings, the plaintiffs allege, could end up in Meta's AI training pipeline without their knowledge.

'No bystanders anywhere had any opportunity to, much less in fact did, consent to having their faces, bodies, voices, and personal information captured, reviewed, labeled, and embedded into Meta's AI systems,' the complaint argues, according to Futurism. Some of the recorded bystanders are minors, and the plaintiffs say 'millions' of people may have been affected.

The litigation dates back to February 2026, when Swedish media reported — based on whistleblowers in Kenya — that footage from the glasses' AI features was being reviewed by human annotators at Sama, a longtime Meta contractor. Workers described being shown highly intimate material, such as people using the toilet, undressing or having sex, and said the automated blurring of faces did not always work. Meta ended its cooperation with Sama weeks after the reports appeared; the outsourcing firm said more than 1,100 employees lost their jobs as a result.

The core allegation has remained the same since the first suits were filed in March: that Meta marketed the glasses as privacy-friendly although recordings made while using the AI assistant were transmitted to the company and partly evaluated by human contractors to improve Meta's AI systems. The plaintiffs are seeking damages as well as court orders that would force changes to Meta's business practices to prevent consumer deception, surveillance of bystanders and privacy violations.

For Meta, the bystander class is particularly delicate: these claimants never accepted any terms of service, so the question of what constitutes effective consent moves to centre stage. Meta rejects the allegations. 'We disagree with these allegations and will fight them,' the company told Futurism, adding that data from its AI assistant may be reviewed 'to help improve our products and people's experiences' — the same practice it says many other companies follow — and that it takes steps to filter identifying information. On the specific situation of recorded bystanders, Meta did not comment.

The case is still at an early stage: Meta has not yet responded substantively to the consolidated complaint, and the court has not ruled on class certification or the merits. But the proceedings are moving — the court set plaintiffs' leadership in July and issued a protocol for future attorneys' fees at the end of that month.

The stakes reach far beyond one lawsuit. Meta reported selling more than seven million pairs of its smart glasses in 2025 alone, and the category is heating up: Google is working with Warby Parker on its own AI glasses, and Apple is reportedly adding cameras and AI to future AirPods. At the same time, societal resistance is growing — courts and cinemas have restricted the glasses, and Meta recently locked down the cameras of devices whose recording LED had been tampered with. If the bystander class is certified, the case could become one of the largest privacy class actions yet against a wearable-AI maker, and a precedent for who is allowed to be in the frame when everyone is recording.

Sources