Kenyan Content Moderators vs Meta — 140+ Former Facebook Workers Diagnosed with PTSD (2023)

Incident Summary

Over 140 former Facebook content moderators based in Nairobi, Kenya, were diagnosed with post-traumatic stress disorder (PTSD) after years of exposure to extreme content including necrophilia, child sexual abuse, and terrorism as part of Meta’s content moderation pipeline.^[1] The moderators, employed through outsourcing partner Sama (formerly Samasource), were paid approximately $1.50 per hour to review and remove the most disturbing content posted to Facebook — a fraction of what US-based moderators receive for similar work.^[2] Non-disclosure agreements (NDAs) prevented the moderators from discussing the nature of their work with family, friends, or external mental health providers, effectively isolating them from the support systems that might have mitigated psychological harm.^[2] The case, which has been proceeding through Kenyan courts, saw its ruling postponed to 2026, extending the workers’ wait for legal resolution.^[3] The incident highlights the human cost of AI content moderation systems that rely on low-paid workers in the Global South to perform the most psychologically damaging work required to make AI platforms safe for users.

Key Facts

• PTSD diagnoses: 140+ former Facebook moderators in Nairobi^[1]
• Pay: Approximately $1.50 per hour^[2]
• Content exposure: Necrophilia, child abuse, terrorism content^[1]
• NDAs: Workers prevented from discussing their work externally^[2]
• Outsourcing: Employed through Sama (formerly Samasource), not directly by Meta
• Court status: Ruling postponed to 2026^[3]

Threat Patterns Involved

Primary: Automation-Induced Job Degradation — The content moderation pipeline demonstrates how AI platforms create degraded labor categories where workers absorb the psychological costs of making AI systems function. The $1.50/hour wages, extreme content exposure, and NDA-enforced isolation represent the human cost that AI content moderation offloads onto workers in the Global South.

Significance

1. Global South labor exploitation — The $1.50/hour wages paid to Kenyan moderators performing the same psychologically harmful work as higher-paid US counterparts demonstrates how AI platforms exploit geographic wage differentials to reduce the cost of content safety
2. NDA as harm amplifier — Non-disclosure agreements that prevented workers from discussing their work effectively blocked access to external mental health support, transforming a workplace hazard into a compounding psychological injury
3. Scale of harm — 140+ PTSD diagnoses among moderators at a single location suggests that the true global scale of moderator harm across all Meta outsourcing locations is substantially larger
4. AI moderation gap — The continued reliance on human moderators for the most extreme content reveals the limitations of AI content moderation systems, which can handle routine violations but still require human exposure to the most psychologically damaging material