Meta Lawsuit Adds Bystanders Who Say AI Glasses Recorded Them Without Consent

Meta’s AI-glasses lawsuit now includes people who never bought the product. An amended federal complaint adds bystanders who allege the glasses captured their faces, voices, bodies and personal information without informed consent, potentially placing that material in a human-review and AI-training pipeline.

The claimed harm moves beyond the customer

Clarkson Law Firm filed the amended class-action complaint in federal court in San Francisco over Meta’s Ray-Ban AI glasses. The proposed class now covers alleged nonconsensual recordings of household members, strangers and minors, rather than only people who used the glasses themselves. The case is in the Northern District of California, and no merits ruling has been issued.

The complaint alleges that potentially millions of people were affected, including minors. It cites alleged recordings involving sexual encounters, bathing children, diaper changes, breastfeeding, showering and changing clothes. Those claims remain allegations, not court findings.

The dispute turns on review, not just recording

The original complaint was filed on March 4 and accused Meta of promoting the glasses as “designed for privacy” while allegedly sending captured audio and video into a human-review process without purchasers’ knowledge. It followed reporting by Swedish newspapers, based on accounts from Kenyan whistleblowers working for Meta contractor Sama, that content captured by the glasses had been stored and sent to human reviewers for AI-training purposes.

The amendment changes the practical boundary of that accusation. Buyers can at least encounter product disclosures; bystanders, the plaintiffs argue, did not own the device or provide informed consent. The filing alleges that captured faces, voices and bodies were labeled and incorporated into Meta’s AI training systems.

Meta rejects the accusation but describes a narrower practice

Meta told Futurism it disagrees with the allegations and will fight them. The company said its glasses enable hands-free AI questions about the surrounding world; if someone uses Meta AI, Meta may review that data to improve its products and people’s experiences. It said the practice works similarly to that of many other companies and that it filters data to help remove identifying information.

That response does not accept the complaint’s account of undisclosed surveillance or its claim that reviewed material was embedded into Meta’s training systems. It instead identifies filtering as the privacy protection at issue. The litigation has not yet established how the court will assess those competing accounts.

A growing product category faces a non-user problem

Meta reported selling more than seven million pairs of smart glasses in 2025. The new plaintiffs’ theory therefore reaches beyond a dispute over an individual buyer’s expectations: it asks whether privacy promises around an AI wearable can address the people its camera and microphone encounter. A separate German regulatory complaint about Meta’s smart glasses, reported in December 2025, raised related concerns about the product category.