According to a report by SVD, Meta’s Ray-Ban AI Smart Glasses have been sending sensitive recordings of people, including “bank details, sex and naked people,” to outsourced companies to review and annotate.
Former Meta employees stated that faces are automatically blurred in annotation data sent off to these subcontractors
It’s ridiculous to think that blurring faces solves privacy concerns when you think about why security cameras couldn’t be put in restrooms.
There are times when cameras are allowed in private spaces, like changing rooms or locker rooms, because of all the thefts. But this is based on a high crime rate. Also, they explicitly state that there are cameras, and it’s easy to check if they’re there before you go, so you can choose not to use them.
But smart glasses are a whole other story. They’re way more invasive and cause a lot more privacy concerns, even though they’re not that good at predicting crime. While users can choose not to use them, companies often deceive users by obscuring what data is collected or burying it within voluminous legal documents. Either way, smart glasses should make sure they’ve got clear consent before sending out video data.
The part where they were asked to send data to meta and it just sends recordings anyway just baffles me. What’s even the point of that toggle, that breaks the whole concept of consent.