Meta Smart Glasses Now Offer AI Visual Data Opt-Out

Meta has recently implemented a significant update to the privacy settings of its smart glasses, now offering users the ability to prevent their visual data from being used for AI training. This development comes amid increasing scrutiny and public discourse regarding data privacy, particularly concerning devices capable of recording and processing real-world interactions. The company's move reflects an ongoing effort to balance technological advancement with user control and trust.
The adjustment to Meta's privacy policy, initially highlighted by Engadget and subsequently confirmed by a Meta spokesperson, marks a crucial shift. Previously, visual data captured by the smart glasses during the use of AI features could be utilized for system training. However, users can now actively choose to opt out of this data collection. This option is accessible through the Meta AI application, where individuals can navigate to their glasses settings, locate the 'Meta AI' section under 'Experiences,' and disable the 'Store visual data from AI experiences' function.
It's important to differentiate this from the standard photo and video capture. Meta has consistently stated that images and videos taken with the smart glasses for personal use remain private and are not employed in AI training by default. The recent policy update specifically targets the visual data generated when interacting with the glasses' artificial intelligence functionalities. This distinction is vital for understanding the scope of the privacy enhancements.
The decision to introduce this opt-out feature follows earlier reports that raised privacy concerns. Notably, a Swedish news outlet revealed that human data annotators in Kenya might have been able to view sensitive visual data, including recordings from intimate moments, captured by Meta's AI glasses. Such revelations underscore the necessity for robust privacy controls in advanced wearable technology. Meta, in collaboration with EssilorLuxottica, the parent company of Ray-Ban, has sold over seven million units of these smart glasses, indicating a substantial user base affected by these privacy adjustments. Furthermore, at the Meta Connect 2026 conference, CEO Mark Zuckerberg announced upcoming models of the glasses that will exclusively feature voice-only AI, devoid of cameras, signalling a diversified approach to product development in response to privacy debates. The issue of privacy has been a recurring challenge for Meta's AI glasses, especially after instances where users tampered with the recording indicator lights, leading Meta to disable camera functions for thousands of devices.
This latest update emphasizes Meta's commitment to addressing user concerns about data privacy and control, particularly as AI-powered wearables become more prevalent. By providing an explicit opt-out mechanism for visual data used in AI training, Meta aims to foster greater transparency and empower users to make informed choices about their personal information while continuing to innovate in the smart technology space.