Meta faced scrutiny after terminating a contract with AI‑training firm Sama, a move that will cut 1,108 Kenyan workers from the project. The partnership fed video from Meta’s smart glasses into pipelines where annotators labeled scenes—including intimate moments—to improve image‑recognition models. Workers who described seeing “naked bodies” said the task was ambiguous and distressing, prompting media investigations in Sweden.

Swedish outlets Svenska Dagbladet and Göteborgs‑Posten published the first accounts, triggering inquiries from the UK’s ICO and Kenya’s Data Protection Commissioner. Meta told the BBC it paused the work to investigate “concerning” claims, emphasizing that users consent to any human review. Sama rejected the breach claim, insisting it met all operational, security and quality standards throughout the engagement.

Activists such as Naftali Wambalo argue the termination was retaliation for staff who spoke out, a charge Meta has not directly addressed. The episode revives debate over privacy safeguards in wearables that can record audio‑visual data in private spaces. As regulators probe the practice, companies will need clearer policies to protect both users and annotators. Failure to address them could erode trust in the platform.