AI data work gets mapped

Brown University has published a story about Cunyan Ma, a student who built an interactive map of the people and companies behind AI data work. It connects data-labeling firms and their corporate customers to worker sites around the world, making visible a part of AI production that businesses and users rarely see. The story drew attention on Hacker News, with 158 points and 138 comments.

The map currently links 13 data-labeling companies, 141 corporate customers and 31 countries with data-worker sites. It isn’t a complete picture, and Ma says it’s meant to be illustrative. The underlying work can include labeling images, rating chatbot responses and reviewing disturbing material. Brown’s report describes companies outsourcing much of this work to lower-cost labor in the Global South.

For a business owner, the practical reminder is that AI doesn’t run on software alone. People do work behind the scenes to prepare and assess the data systems depend on. If your company uses AI, knowing who does that work, and under what conditions, is part of understanding what sits behind the tool. Brown describes Ma’s map as a public education tool for showing the wider economic and political structure behind AI. Read Brown’s report.