Brown student maps invisible labor networks behind AI training

A Brown University student has created an interactive mapping tool that documents the global labor supply chains feeding AI model development. The research highlights how data annotation, labeling, and other training work is systematically outsourced to workers in the Global South at significantly lower wages than equivalent work in wealthy nations.

The tool visualizes the power imbalance in AI development: major tech companies rely on vast networks of contractors and subcontractors across countries like the Philippines, India, Kenya, and others to prepare training data, yet these workers remain largely invisible in public narratives about AI progress. The project exposes wage disparities and working conditions that enable the cost economics of modern AI systems.