Meta in Row After Workers Who Say They Saw Smart Glasses Users Having Sex Lose Jobs

Why It Matters

The fallout underscores the reputational and regulatory risk of using human reviewers for wearable AI data, and it raises questions about labor practices in the emerging AI‑annotation market.

Key Takeaways

• Meta cut ties with Sama, risking 1,108 Kenyan jobs.
• Workers reported reviewing footage of sex and bathroom use from smart glasses.
• UK ICO and Kenyan regulator launched investigations into privacy breaches.
• Meta cited “standards” while Sama disputes any performance failure.
• Glasses’ misuse fuels debate over AI ethics and human oversight.

Pulse Analysis

Meta’s AI‑enabled smart glasses, unveiled in partnership with Ray‑Ban and Oakley, promise real‑time translation and visual assistance for users, especially those with visual impairments. The devices embed a tiny camera that activates a red light when recording, and Meta’s terms of service disclose that captured content may be reviewed by humans to improve algorithms. While the technology offers compelling use cases, the reliance on human annotators to label and moderate live footage introduces a privacy frontier that many regulators have yet to fully address.

In February, workers at the Kenyan contractor Sama claimed they were asked to examine videos showing intimate moments and bathroom activities recorded by the glasses. After the allegations reached Swedish media, Meta paused and then ended its partnership with Sama, citing unmet standards. Sama, a certified B‑Corp, disputes any performance breach and blames the termination on employee retaliation. The contract’s termination threatens over a thousand jobs and has prompted the UK’s Information Commissioner’s Office and Kenya’s Data Protection Commissioner to launch formal inquiries into potential breaches of data‑protection law.

The controversy highlights a broader industry dilemma: balancing the benefits of wearable AI with robust ethical safeguards. As more firms explore head‑mounted devices for consumer and enterprise applications, transparent consent mechanisms, independent oversight, and clear standards for human review will become essential to avoid regulatory backlash and public mistrust. Meta’s handling of the incident will likely influence how other tech giants structure their data‑annotation pipelines and could accelerate calls for stricter global guidelines on AI‑driven surveillance tools.