Meta Platforms has been sued in California over claims that its AI glasses recorded private conversations without the wearer ever saying the “Hey Meta” wake phrase. The proposed class action, filed on Monday in the US District Court for the Northern District of California, alleges that the glasses captured conversations — including sensitive discussions with doctors and financial institutions — without the knowledge or consent of the people being recorded.
The plaintiff, Angelo Subia, says he bought a pair of Meta AI Glasses at a LensCrafters store. According to the complaint, the glasses recorded his conversations even when he never used the trigger phrase designed to activate the device, and the recordings were allegedly made in real time as the conversations happened.
What The Complaint Alleges
The case is Subia v. Meta Platforms Inc., No. 3:26-cv-12135. Subia seeks to represent a class of all California residents whose conversations were recorded by the glasses without the trigger phrase. The complaint says users discovered the problem by checking their Voice Activity Logs inside the companion app, and that some users noticed the glasses’ recording indicator light activating after a firmware update.
The claims are brought under the California Invasion of Privacy Act, which bars recording confidential communications without the consent of all parties. Subia alleges violations of the act’s wiretapping and eavesdropping provisions and is seeking compensatory and statutory damages, injunctive relief, attorneys’ fees and costs, and pre-judgment interest. Bursor & Fisher PA represents Subia and the proposed class. Meta did not respond immediately to a request for comment, Bloomberg Law reported, and the allegations in the complaint have not been tested in court.
How The Wake Phrase Is Supposed To Work
Meta’s glasses are built around a voice-first design: the wearer says “Hey Meta” to wake the assistant, ask questions, take photos and video hands-free, and hear responses through the frame’s speakers. The recording indicator light is the outward signal to people nearby that the device is capturing. Our explainer on Meta AI and its glasses sets out how the assistant, the Llama models behind it and the eyewear fit together as one product line.
The lawsuit turns on that signal chain. If recordings can begin without the wake phrase — whether through a fault, a firmware change or some other mechanism — then neither the wearer’s intent nor the indicator light reliably tells the people around the glasses that a conversation is being captured. The complaint’s reference to Voice Activity Logs matters for the same reason: the logs are the record wearers themselves would check to learn what the device heard, and the suit says those logs are where the unexpected recordings surfaced.
A Fast-Growing Product Meets A Consent Problem
The complaint says Meta sold more than 7 million camera-equipped pairs of the glasses in the past year, a scale figure that underlines how quickly camera-and-microphone eyewear has moved from novelty to mass product. Meta holds a commanding share of the AI glasses market, built on the Ray-Ban partnership and distribution through ordinary eyewear stores, which is precisely why a consent defect alleged at this scale draws a class action rather than an individual complaint.
The suit also lands in a run of privacy pressure on Meta’s data practices. This month the company first expanded, then removed, AI training language from its commercial terms after publisher pushback, and its business terms had already drawn scrutiny for claiming a right to crawl business websites from October 30. Those disputes concerned web data; the glasses case concerns the physical world, where the people recorded may have no relationship with Meta at all.
Why It Matters
The Meta glasses lawsuit matters because it tests the consent model that always-on wearables depend on. A phone records when its owner opens an app; glasses sit on a face in rooms full of other people, and the wake phrase plus indicator light are the entire mechanism by which those other people are meant to be protected. If a court finds that recordings happened outside that mechanism, the exposure is not limited to one plaintiff: the proposed class covers California residents recorded without the trigger phrase, and California’s privacy act carries statutory damages per violation. For rival wearable makers, the case is a warning that the audit trail — logs, indicator behaviour, firmware changes — is now part of the product’s legal surface, not just its engineering. And for Meta, whose glasses are its main hardware route for the assistant, a consent fight over recording strikes at the feature the product is sold on.
Background
Meta’s AI glasses grew out of its Ray-Ban partnership, which turned camera eyewear into a mainstream product line and gave the Meta AI assistant a wearable home alongside WhatsApp, Instagram and Facebook. Privacy questions have followed camera glasses since the category began, and regulators and plaintiffs have repeatedly focused on bystanders who never agreed to be recorded. This suit is the latest and most specific version of that argument: not that the glasses can record, but that they allegedly did so when they were not supposed to.
Source: Bloomberg Law; Law360
0 Comments