The AI industry's biggest open secret is its dependence on a vast, largely invisible workforce of data laborers — millions of workers across the Global South who label images, transcribe audio, moderate toxic content, and provide the human feedback that makes AI systems function. Without them, systems like ChatGPT, Claude, and Gemini would be unable to distinguish helpful responses from harmful ones. Yet these workers are among the most exploited in the technology supply chain (Brookings, 2025).
The wage disparities are staggering. OpenAI contracted with outsourcing firm Sama, paying the firm $12.50 per hour while workers in Kenya received just $2 per hour for some of the most psychologically damaging work imaginable: reviewing and labeling content depicting graphic violence, sexual abuse, and hate speech. Their American counterparts performing similar work earn $20 per hour — a 10x pay gap for identical tasks (Time Magazine, 2023; Final Call, 2024).
The psychological toll is devastating. A 2025 survey by Equidem of 76 workers from Colombia, Ghana, and Kenya reported 60 independent incidents of psychological harm, including anxiety, depression, panic attacks, PTSD, and substance dependence. Over 140 workers in Kenya alone were diagnosed with severe PTSD. Workers described being haunted by images of child abuse, torture, and extreme violence that they had to view and categorize for hours daily — with little or no access to mental health support (Equidem, 2025).
The exploitation is structural. Data workers are typically hired as independent contractors through layers of subcontractors, denying them employee protections, benefits, or job security. One investigation revealed that Kenyan data labelers working for the platform Remotasks were unaware it was a subsidiary of ScaleAI, a company valued at over $13 billion. In March 2024, Remotasks abruptly shut down its Kenya operations, leaving thousands without income or warning (LSE, 2025; GIJN, 2025).
Researchers describe this pattern as "AI colonialism" — a new form of exploitation where wealthy nations and corporations extract labor from the Global South to build technologies that primarily benefit the Global North. Workers have begun organizing: Kenya's Data Labelers Association was launched to fight for better pay, mental health support, and transparency. But without international regulation and corporate accountability, the hidden workforce of AI will continue to bear the human cost of artificial intelligence (JHU, 2025; RFK Human Rights, 2024).
Key Sources
- Brookings Institution (2025). Reimagining the future of data and AI labor in the Global South.
- Equidem (2025). Survey of data workers in Colombia, Ghana, and Kenya.
- LSE Media (2025). The perilous future of AI work in the Global South.
- Project MUSE / JHU (2025). Artificial Intelligence Colonialism.