Images and videos processed and analysed through Ray-Ban Meta smart glasses are being used to train Meta's AI systems, prompting concerns that users may inadvertently be sharing sensitive information with the technology company. 

Meta's privacy policy states that users' content - which includes potentially sensitive information captured about other people - shared with the company in regions where multimodal AI is available (currently the US and Canada) could contribute to enhancing its AI capabilities.

While photos and videos captured by the glasses without user submission are not used for training, once a user requests analysis from Meta AI, those images fall under different privacy considerations. This can lead to potential privacy and confidentiality issues, as users might unknowingly share sensitive personal content, including images of their homes or loved ones. 

Meta's recent updates have introduced new features for its smart glasses, allowing users to engage with AI more naturally, which may increase the likelihood of sharing personal data for analysis. 

The company previously trained its AI models using publicly available data from platforms like Facebook and Instagram, but this extension to smart glasses raises concerns about user awareness and consent regarding the data being collected and used for AI training.