Meta has confirmed that it may utilize any images users request Ray-Ban Meta AI to analyze for training its artificial intelligence systems. Initially, the company was reticent to provide details on this matter, but further clarification has emerged. According to Emil Vazquez, a policy communications manager at Meta, images and videos shared with Meta AI in regions where multimodal AI is available, such as the U.S. and Canada, can be used to enhance the AI's capabilities in accordance with their Privacy Policy. The distinction is made that while photos and videos taken with Ray-Ban Meta smart glasses are not used for training unless submitted to the AI, once a user requests an analysis, those images are subject to different policies. This means that users inadvertently contribute to a growing database that Meta can leverage to refine its AI models. The only way to avoid this is to refrain from using the AI features altogether. This situation raises significant privacy concerns, as users may not fully grasp that they are providing Meta with potentially sensitive images, which could include personal spaces or identifiable individuals. Although Meta asserts that this process is made clear in the user interface, there seems to have been a lack of initial transparency from the company regarding these practices. Previously, it was known that Meta trains its Llama AI models on publicly available data from platforms like Instagram and Facebook, but this new approach extends that definition to include any images analyzed through the smart glasses. The timing of this revelation is particularly pertinent, as Meta has recently introduced new AI features that encourage users to interact with the AI in a more intuitive manner, increasing the likelihood of sharing new data for training purposes. A notable addition is a live video analysis feature that streams images to Meta’s AI models, allowing users to receive outfit suggestions based on their wardrobe. However, the company does not prominently disclose that these images are also being sent to Meta for training purposes. Meta's privacy policy explicitly states that interactions with AI features can be utilized for training AI models, which encompasses images shared through the Ray-Ban smart glasses. Furthermore, the terms of service indicate that by sharing images, users consent to Meta analyzing those images, including facial features. This situation is compounded by Meta's recent legal history, having settled a $1.4 billion lawsuit in Texas concerning its facial recognition practices. The case revolved around a feature called "Tag Suggestions" on Facebook, which was made opt-in after significant backlash. Notably, some of Meta AI's image features are not available in Texas due to these legal issues. Additionally, Meta retains transcriptions of voice interactions with Ray-Ban Meta by default to train future AI models, although users can opt out of having their voice recordings used for this purpose when they first log into the app. The broader context involves a trend among tech companies, including Meta and Snap, to promote smart glasses as a new computing platform. These devices, equipped with cameras, raise privacy concerns reminiscent of the issues surrounding Google Glass. Reports have surfaced of individuals hacking Ray-Ban Meta glasses to access personal information about those they encounter, further highlighting the potential risks associated with this technology.