Meta has announced that users of its AI-powered smart glasses can now choose not to have their "visual data" used for training artificial intelligence models. This update comes in response to growing privacy concerns about how the company collects and uses data from its wearable devices. Under the new setting, users who opt out will have their visual data used only for real-time assistance from Meta AI—such as answering questions—and then the data will be deleted immediately. It will not be stored, used for AI training, or reviewed by third parties. Previously, images and videos captured during interactions with Meta AI were used to improve the AI models, even if they were not saved to a user's personal camera roll. Meta has clarified that the actual content of photos taken by the glasses' cameras is not shared or used for AI training. However, this distinction was not clearly explained to users in the past, leading to confusion and concern. Reports also suggest that some third-party contractors in other countries have reviewed images captured by the glasses, including content that could be considered personal or intimate. The new privacy setting applies only to visual data, meaning audio recordings from interactions with Meta AI are still available for training and review. However, users can manually delete these audio recordings if they choose. In an effort to provide more privacy options, Meta has also launched a new line of Ray-Ban-branded smart glasses that do not include cameras. This move offers users an alternative if they prefer not to have any visual data collected at all.