Two United States senators have demanded answers from Meta about privacy concerns linked to its smart glasses, warning that people who have not agreed to be recorded could have their images captured and used to train artificial intelligence systems. Senators Ron Wyden of Oregon and Edward Markey of Massachusetts sent the letter to Meta on Wednesday, asking what safeguards exist for bystanders and what data the company uses for AI training. The letter seeks a detailed response on practices that, the senators argue, raise serious questions about consent.
What the senators asked
The senators wrote that Meta’s own disclosures confirm imagery processed through the glasses’ visual AI features may be retained and used to train AI systems, including with human reviewers. Using intimate imagery for AI development without consent is itself an invasion of privacy, the letter argues, especially for bystanders. The concern is not only recording, but later use.
Among their questions, the senators asked what safeguards prevent the nonconsensual capture, disclosure or processing of sensitive and intimate imagery. They also asked whether Meta evaluated whether the glasses could facilitate voyeurism or nonconsensual intimate image capture. Those questions go to product design and risk assessment, asking whether harms were considered before the devices reached the public.
The letter asks what imagery Meta uses for training, and whether people can request deletion of data captured by the glasses. Deletion matters because a briefly captured recording may have a long life inside training datasets, review queues or derived models. The senators seek clarity on how long such data is kept and who can have it removed.
Reports cited in the letter
The senators’ concerns follow a series of earlier reports about smart glasses and privacy. An October 2025 investigation examined people using Ray-Ban glasses to harass massage-parlour workers, raising questions about how wearable cameras can be misused in private or semi-private settings. The letter places that reporting in a broader pattern, rather than treating any single incident as isolated.
A January BBC report cited a woman recorded with the glasses who said she was bombarded with calls after her number was posted. The letter also refers to more than 70 people who sued Meta this year, claiming intimate images and videos reached data labellers. Those claims underline why human review of captured material is a central concern.
Swedish journalists reported in March that Kenyan labellers viewed uncensored smart-glasses footage, and Meta ended that contract less than two months later. Together, the senators argue, these episodes show a gap between hands-free technology’s promise and protections for people appearing in its footage without consent.
What Meta has been asked to explain
Meta has not yet responded in the material read. The letter seeks answers, not a ruling, inviting the company to explain its safeguards, training practices and deletion options. Any reply will be an important next step in assessing how the company views bystander privacy.
The questions touch Meta’s wider AI work, since the glasses’ visual features feed into systems that learn from imagery. As Meta expands its AI assistant and products, the senators’ letter asks whether consent and privacy protections are keeping pace with deployment. The distinction between willing users and unknowing bystanders sits at its heart.
Why It Matters
Smart glasses change the usual privacy calculation because recording can be discreet, continuous and hands-free. A bystander may not realise a device is capturing them, may have no relationship with the wearer, and may have no obvious way to object at the time. That makes consent and deletion harder to request afterwards than with a phone held visibly in hand.
The training question adds a second layer. Even if an image is not published, its use to train an AI system, especially with human reviewers involved, can feel like a further intrusion for the person pictured. Lawmakers are testing whether privacy practices are adequate for wearable AI, or whether new safeguards, clearer disclosure and stronger deletion rights are needed.
Background
Smart glasses combine a camera, audio and AI features in eyewear designed for everyday use, and their growth has brought wearable privacy questions that were once limited to phones and security cameras into streets, workplaces and private businesses, raising new questions about consent, disclosure and deletion.
0 Comments