Summary
The UK's data watchdog is asking Meta for information about claims that workers viewed private videos from Meta's AI glasses. Some videos included sensitive content, and Meta said these were reviewed by workers to improve their technology. The workers review content to teach the AI system but say there are privacy issues with what they see.
Key Facts
- The UK's Information Commissioner's Office is contacting Meta about reports of workers viewing sensitive content from AI glasses.
- Meta uses subcontracted workers to review videos captured by its AI smart glasses to enhance user experience.
- Some sensitive videos, like those of people in private moments, were reportedly seen by workers in a Kenya-based outsourcing company.
- Meta claims it has privacy measures, such as blurring faces, but reports indicate these do not always work.
- Users must manually activate recording on the smart glasses, but they may be unaware their content can be reviewed by people.
- The outsourcing company in Nairobi, Sama, employs workers who help teach Meta's AI by labeling content.
- Sama workers check AI responses for accuracy and have strict privacy rules at their workplace.
- Meta launched a range of AI-powered glasses in partnership with brands Ray-Ban and Oakley in September 2025.
