Brown student's AI labor map links 13 firms, 31 countries

An undergraduate at Brown University has published an interactive map that connects 13 data labeling companies to 141 corporate customers and plots data worker sites in 31 countries. The map, at ai-materiality-map.org, is the work of Cunyan Ma, a senior concentrator in Social Analysis and Research, Brown reports.

Ma built the project through the University's Center for Technological Responsibility, Reimagination and Redesign, with computer science professor Suresh Venkatasubramanian advising. It aggregates existing databases, Ma's own research and news reports on data labor into a single view of workers who do image labeling, speech transcription, chatbot response rating, content moderation, and tasks like identifying tumor boundaries in medical scans.

"My primary motivation is to visualize the idea of AI colonialism in an impactful and immediately accessible manner," Ma says. "The gap I'm trying to fill is that most of what we know about data workers comes from anecdotal reports."

Some of what the map covers is grim. Describing the labor behind AI safety training, Ma says workers have to read "paragraph after paragraph of AI-produced child pornography, bestiality, violent content." The sites cluster in Kenya, Indonesia, Colombia and other Global South countries where firms seek inexpensive labor.

Venkatasubramanian puts the point plainly: "It's all humans underneath." That, he says, is "the real story behind the generation of AI."

The 13 vendors and the 141 customers are not individually listed in the Brown write-up, though they are described as ranging from "social media and tech firms to retailers and defense companies." Ma plans to extend the project next into power maps of the semiconductor industry's labor.