Meta AI Glasses Human Review Footage: Videos recorded from Meta AI smart glasses are being reviewed and annotated by data workers in Kenya for AI training. Reports claim that some of the clips included very private moments.
Meta AI Smart Glasses Privacy Controversy: Have you ever wondered where the video recorded from the AI smart glasses you wear goes? If you are using Meta’s AI smart glasses, then this news is very important for you. According to reports, data workers sitting in Nairobi, Kenya are watching and tagging the videos that users have recorded with Meta Ray-Ban Smart Glasses. These clips are being used in the ‘data labeling’ process for AI model training. Let us know what is the whole matter and why it is considered dangerous…
what is the whole matter
According to reports by Swedish newspapers Goteborgs-Posten and Svenska Dagbladet, some Kenyan employees claimed to have seen videos in which people were seen going to the toilet, changing clothes or having private moments. The work of processing these videos is with the tech contractor company ‘Sama’, where ‘data annotators’ have to tag them after watching the videos. An employee was quoted as saying, ‘We see everything from living rooms to nude bodies. In some videos, people are going to the toilet or taking off their clothes. Perhaps they don’t even know that they are being recorded.
What does Meta’s privacy policy say?
Meta claims that their smart glasses have been designed keeping privacy in mind and users have control over their data. But their policy clearly states that ‘In some cases Meta may review your AI interactions, including conversations or messages. This review can be automated or manual (human). It has also been warned not to share sensitive information which AI can store or use. Through the ‘Hey Meta’ command in the glasses, the user can turn on the camera, make a video call or record a photo or video.
How many people have these glasses reached?
Sales of these AI glasses, created in collaboration with eyewear giant EssilorLuxottica, have increased rapidly. A total of 20 lakh units have been sold in 2023-24 and 70 lakh units in 2025 alone. Meaning, millions of people are using them in everyday life.
Why is this matter very dangerous?
According to experts, data labeling is a necessary process for AI training, but are users aware of this level of manual review? Do they know that someone thousands of kilometers away can view their recorded moments? What can happen if someone’s bank card or personal conversation or private moments are inadvertently captured on camera? If the data gets leaked, it could become an ‘enormous scandal’, the employees themselves have feared.
Read this also-Online Scam Alert: Catch fake website in just 1 minute, know the simplest tricks
Read this also-WhatsApp Tips: If someone has blocked you, find out this way with Meta AI