When Google launched its Glass headset in 2013, the backlash was swift. Wearers were dubbed “Glassholes” and banned from bars, restaurants and cinemas. The concern then was largely around privacy, the discomfort of not knowing whether someone was recording you. Google eventually shelved the consumer version in 2015. A decade later, the smart glasses debate is back with Meta’s Ray-Ban glasses, only this time the privacy implications run considerably deeper.
About 7 million pairs of the Ray-Ban smart glasses have been sold in 2025. Marketed as an AI assistant that processes what the wearer sees, the product has found a genuine consumer audience. Unlike traditional cameras, these glasses process real-time visual and audio data to power AI assistants.
But a joint investigation published in late February by Swedish newspapers Svenska Dagbladet and Göteborgs-Posten has raised serious questions about what actually happens to that footage. This investigation found that footage captured through the glasses, once users activate the AI assistant, is routed to a Kenyan data annotation firm subcontracted by Meta to label and categorise content for AI training. Workers described viewing footage of people undressing, credit card details, and other personal and intimate content, often captured without the wearer’s awareness that their camera was active. And almost never with the knowledge of the person being filmed that a stranger was watching them.
Meta’s own terms of service include language permitting manual (human) review of AI interactions.
The situation was compounded in April 2025, when Meta updated its privacy policy, enabling AI features by default and removing the option for users to prevent voice recordings from being stored. While Meta maintains that human review is essential for training AI models and that the data is anonymised, whistleblowers claim the blurring of faces is inconsistent.
The devices have drawn separate scrutiny for how they are used in public spaces. Content creators have used the technology to film strangers for social media, often without knowledge or consent. There have also been several instances across the world of male influencers using the Meta Ray-Ban glasses, among smart glasses of other companies, to secretly film women in public, posting the footage on TikTok and Instagram as dating advice content. Victims have faced mass online harassment, with phone numbers and personal details inadvertently captured and leaked. In October 2025, the University of San Francisco issued a warning after reports that a man wearing Ray-Ban glasses was filming women and posting the footage online.
Reports from February 2026 suggest Meta now plans to integrate facial recognition features to its smart glasses—a move that would allow wearers to identify strangers in real-time by cross-referencing social media databases, effectively ending public anonymity.
But this is not the first instance where Meta’s wearable technology has upset the law. In February, while appearing in a Los Angeles Superior Court to testify in a trial regarding social media addiction, Meta CEO Mark Zuckerberg’s team was reported to have been wearing the Ray-Ban glasses.This prompted the judge presiding over the trial to issue a stern warning after noticing the glasses, “If you have [recorded proceedings], you must delete that, or you will be held in contempt.”
In India, under the “Consent and Transparency” framework of the Digital Personal Data Protection (DPDP) Act, 2023, manufacturers must obtain explicit consent for processing data, restrict AI training to disclosed purposes and grant users the right to data erasure once primary goals are met. Violations under the DPDP Act carry penalties up to Rs250 crore.
Positioned as a tool for India’s mobile workforce, Kaze targets field operations, inspections and data collection, alongside education, health care and assistive use for the visually impaired. The company has also flagged a potential role in India’s fully digital Census 2027.
First Published: Mar 05, 2026, 16:49
Subscribe Now