Meta’s AI glasses reportedly send sensitive footage to human reviewers in Kenya

A major privacy breach has been exposed involving Meta's popular AI-powered smart glasses. An investigation by Swedish outlets Svenska Dagbladet and Göteborgs-Posten alleges that sensitive video footage captured by users' Ray-Ban and Oakley smart glasses is being sent to human reviewers in Nairobi, Kenya. According to the report, AI annotators contracted by Meta have viewed footage containing bathroom visits, naked individuals, and intimate moments, directly contradicting Meta's marketing that the glasses were "designed to protect privacy." This has already triggered at least one proposed class action lawsuit in the US, accusing Meta of false advertising and privacy violations. The UK's Information Commissioner's Office is also questioning the company about these claims.

The technical details reveal a significant gap between policy and practice. While a former Meta employee stated that faces in annotation data are automatically blurred, Kenyan contractors reported this system "does not always work as intended," leaving identifiable people and even bank card details visible. The glasses, which sold over 7 million units in 2025 through partner EssilorLuxottica, feature a 'Hey Meta' AI assistant that answers questions based on what the camera sees. Meta's spokesperson stated that media "stays on the user's device" unless shared, and that using contractors to review shared data is standard for improving AI. However, privacy advocates warn this incident, coupled with Meta's alleged interest in adding facial recognition, poses a grave risk to civil liberties and consumer trust in wearable AI.