Meta’s smart glasses have a new opt-out.
Meta spokesperson Dina El-Kassaby confirmed to Business Insider that the company is changing the setting options on its camera-wielding glasses.
Previously, the glasses could collect visual data for training when the wearer used AI features. Now, those users can opt out of having their visual data stored for AI training. Engadget first reported the change.
In a blog post, Meta wrote that users can do so in four steps: Open the Meta AI app, tap the glasses settings, find “Meta AI” under “Experiences,” and turn off “Store visual data from AI experiences.”
Meta has long said that it doesn’t save data from ordinary photos and videos taken with the glasses. A user’s camera roll is private by default and not used to train AI, the company said.
That’s different for visual data collected during AI use. The current privacy policy, which has been in effect since June, says Meta can “share recordings, transcripts, and related data about your voice conversations with Meta AI with vendors and service providers who assist us in improving, troubleshooting, and training our speech recognition systems.”
Earlier this year, a Swedish news outlet reported that human data annotators in Kenya could video and images taken while using the AI features of Meta glasses, including during intimate moments.
Meta makes the glasses with EssilorLuxottica, Ray-Ban’s parent company. The brands reported selling more than 7 million units last year. On Wednesday, during the company’s Meta Connect 2026 conference, CEO Mark Zuckerberg announced that two new styles will come with camera-free, voice-only AI features.
Meta’s AI glasses have spawned uneasy questions over privacy. They’re designed to flash a small light to alert bystanders when they’re recording — but Meta recently said it had to deactivate the hardware for thousands of users who had tampered with the lights.
