More than 70 people have sued Meta over its AI-enabled smart glasses, alleging that the devices captured intimate footage and sent material to workers involved in training Meta’s artificial-intelligence systems. The allegations matter because they challenge the gap between a recording light, user consent and what an AI assistant may see when people activate it.
Meta AI Watch analysis
What happened
The Los Angeles Times reports that a proposed class-action lawsuit filed in federal court in Northern California accuses Meta of fraud, false advertising and breaches of consumer-protection law. The plaintiffs allege that the glasses recorded people in private situations, including while undressing, using the bathroom or entering passwords, sometimes without their knowledge.
The lawsuit further alleges that footage and audio captured when users activated Meta AI was sent to Meta’s servers and reviewed by contractors in Kenya who helped label material used to train the company’s AI systems. The paper reports that the complaint includes claims from users who say the glasses activated accidentally or continued recording after being placed down.
Meta disputes the allegations and says it intends to fight the case. A company spokesperson told the Los Angeles Times that Meta may review data from Meta AI to improve its products, while taking steps to filter material, remove identifying information and protect privacy. Meta’s published explanation says the camera activates when a user asks Meta AI about something in front of them, such as a menu or plant, and that faces and number plates are blurred.
Read the Los Angeles Times report
Why it matters
This is a legal complaint, not a court finding that Meta broke the law. But it raises a practical question every wearer and bystander should understand: the privacy risk may begin before a conventional recording is deliberately saved or shared.
Meta’s glasses are designed to make AI interaction feel effortless. That convenience also makes the boundary between an intentional request and an unintended capture unusually important. A light can signal that the glasses are recording, but it cannot by itself explain what data has been sent, who may review it or whether a person nearby has meaningfully consented.
Our read
The strongest part of this story is not the lawsuit’s most lurid allegation. It is the proposed chain from a spoken AI request to camera footage, remote processing and human review. If that chain is accurate, users need clearer controls and explanations than a tiny indicator light can provide.
Meta’s response and its stated blurring safeguards are relevant, but they do not settle the plaintiffs’ claims. The court will have to test what the glasses captured, when they transmitted it, what users were told and how the data was handled. Until then, treat the allegations as allegations, while recognising that the underlying design question is already here: can wearable AI observe responsibly in spaces where people never agreed to become training data?
What to watch
- Whether Meta produces technical records showing when the glasses activate, transmit and retain footage.
- Whether the court allows the proposed class action to proceed.
- What safeguards Meta announces for accidental activation and bystanders.
- Whether regulators or lawmakers address AI-enabled observation separately from ordinary video recording. Sources & Evidence: The Los Angeles Times, 18 September 2026, reports on a proposed class-action lawsuit against Meta. The allegations, quoted user accounts and claims about contractor review are attributed to the complaint and the newspaper’s reporting. Meta disputes the allegations and says it intends to fight the case. The supplied evidence does not establish liability or independently verify the plaintiffs’ accounts.
Discussion spark: Should AI glasses be required to obtain clearer consent for every camera-assisted request, or would that make their everyday usefulness impossible?
Sources and evidence
- Meta glasses captured and shared intimate images without users’ consent, lawsuit claims – Los Angeles Times (18 September 2026, 10:00 UTC)
not affiliated with, endorsed by, or operated by Meta