TL;DR
The UK Information Commissioner’s Office (ICO) is writing to Meta after a BBC investigation revealed that outsourced workers in Kenya viewed sensitive and intimate content captured by Ray-Ban Meta AI smart glasses. Workers at a company called Sama saw footage including bedroom and bathroom recordings. Meta says the content was filtered before review, but a Swedish investigation found face-blurring technology sometimes failed.
What Was Found
Workers employed by Sama, a Kenyan outsourcing firm contracted by Meta, were tasked with reviewing content captured through Ray-Ban Meta AI smart glasses to help improve the product experience. During this review process, workers encountered intimate videos — including footage recorded in bedrooms and toilets.
Meta maintains that content is filtered before it reaches human reviewers. However, a separate investigation conducted in Sweden found that the face-blurring systems applied to this content sometimes failed, meaning workers could see identifiable individuals in private settings.
The scale of the exposure remains unclear. Neither Meta nor Sama have disclosed how many workers had access to this content, how many recordings were reviewed, or how long the practice continued.
The ICO’s Response
The ICO confirmed it is writing to Meta to seek clarity on the company’s data handling practices related to the smart glasses. The regulator stated that devices of this nature “must put users in control” of their data.
While the ICO has not yet opened a formal investigation, the decision to contact Meta directly signals that the regulator considers the reports serious enough to warrant immediate engagement. Under UK GDPR, companies processing personal data must ensure adequate safeguards are in place — including when that processing is outsourced to third-party contractors in other jurisdictions.
Wider Context
Ray-Ban Meta smart glasses have sold well since their launch, offering features including live video capture, AI-assisted visual queries, and hands-free calling. But the privacy implications of always-on camera-equipped eyewear have concerned regulators since Google Glass first raised similar issues a decade ago.
The use of outsourced workers in lower-cost markets to review AI training data and user content has drawn increasing scrutiny. Previous investigations have documented poor working conditions and psychological harm experienced by content moderators at outsourcing firms working for major technology companies.
Looking Forward
The ICO’s engagement with Meta will be watched closely by privacy advocates and the wearable technology industry alike. If the regulator determines that Meta’s data handling fell short of UK standards, it could set a precedent for how AI-powered wearable devices are governed in Britain — and force manufacturers to rethink the role of human review in training consumer AI products.