Workers based in Nairobi, Kenya, are reviewing video footage captured by Meta’s AI-powered smart glasses as part of the company’s efforts to train artificial intelligence systems. This claim comes after a report was published by Swedish newspapers Goteborgs-Posten and Svenska Dagbladet.
The clips come from Meta Ray-Ban smart glasses, wearable devices equipped with built-in cameras and microphones that allow users to record videos, take photos, and interact with an AI assistant. The recorded material is reportedly sent to a technology contractor called Sama, where employees work as data annotators.
Data annotators reviewing user footage
Data annotators are hired to watch and label content so that AI models can better understand images, objects and human behaviour. The process is widely used in the development of artificial intelligence.
Workers involved in the project claim that some of the videos they reviewed appeared to show people in highly private situations. According to the report, clips allegedly included individuals using the toilet, changing clothes or engaging in intimate activities.
“We see everything, from living rooms to naked bodies. There are also sex scenes filmed with the smart glasses, someone is wearing them having sex,” one worker was quoted as saying in the report.
Also Read |MacBook Neo could be Apple’s cheapest laptop yet: Here’s what we know so far
Another worker suggested that many people appearing in the recordings may not realise they are being filmed.
“In some videos you can see someone going to the toilet, or getting undressed. I don't think they know, because if they knew they wouldn't be recording,” the worker said.
Some of the footage reportedly showed sensitive personal information, including bank cards that were visible by mistake.
Workers describe pressure to continue reviewing
One employee said the job required them to keep reviewing and labelling videos even when the content felt intrusive.
“You understand that it is someone's private life you are looking at, but at the same time you are just expected to carry out the work,” the worker said, adding that questioning the process could risk their job.
What Meta’s privacy policy says
Meta has said the smart glasses are designed with privacy controls and that users manage their own data. However, the company’s privacy policy notes that interactions with its AI systems may be reviewed.
Also Read | Fortnite finally coming back to Play Store after years-long Epic vs Google battle
“In some cases, Meta will review your interactions with AIs, including the content of your conversations with or messages to AIs, and this review may be automated or manual (human),” the policy states.
“Do not share information that you don't want the AIs to use and retain, such as information about sensitive topics”, it further states.
The glasses are activated using the voice command “Hey Meta”, allowing users to take photos, record videos or make first-person video calls.
Manufactured with eyewear company EssilorLuxottica, Meta sold around seven million units of its AI glasses in 2025, after selling two million devices across 2023 and 2024 combined.
