Meta is enhancing its AI capabilities by leveraging data collected from Ray-Ban Meta smart glasses. A recent privacy policy update reveals expanded data collection practices for training AI models, raising privacy concerns.
With the “Hey Meta” feature active, images captured through the glasses’ camera can be used for AI training. While Meta assures users the camera doesn’t constantly record, any photos or videos taken are fair game. This always-on nature of the “Hey Meta” feature, designed for seamless interaction with the AI assistant, inevitably raises questions about continuous data collection.
Furthermore, Meta has removed the option to prevent voice recordings from being stored on its servers. Users must now manually delete each recording to limit Meta’s access. The company’s voice privacy notice states that voice transcripts and audio recordings are stored for up to a year to improve Meta’s products. Accidental voice interactions are retained for 90 days.
These changes come as Meta aggressively expands its AI initiatives. The company recently launched a live translation feature on the Ray-Ban smart glasses and a standalone Meta AI app. This push for AI development underscores Meta’s need for extensive data to refine its models, particularly after allegations of manipulating benchmark test results.
This trend towards increased data collection from devices with microphones and cameras raises broader concerns about privacy. As companies prioritize the value of captured data over user privacy, devices like smart glasses and home speakers could increasingly function as surveillance tools. The trade-off between convenience and privacy remains a critical consideration for users in an increasingly AI-driven world.
Meta was contacted for comment on these policy changes, but no response was received by the time of publication.
