“Ghost work” is the term anthropologist Mary L. Gray and computer scientist Siddharth Suri introduced in their 2019 book “Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass,” published by Houghton Mifflin Harcourt. It names the large, mostly invisible human labor force that keeps supposedly automated systems running. As the book’s site puts it, “work isn’t disappearing in the age of AI; it is being hidden.” The publisher’s description states that services from “Amazon, Google, Microsoft, and Uber can only function smoothly thanks to the judgment and experience of a vast, invisible human labor force.”
The work itself is high-tech piecework: flagging explicit content, labeling and proofreading data, transcribing audio, and patching the gaps where algorithms fail. Drawing on a five-year study of workers in the United States and India, the authors estimate that around 8 percent of Americans have done such work, and that the platforms typically leave these workers earning below traditional minimums, without benefits, and dismissable “at any time for any reason, or none.” The book ties this directly to AI: much of what looks like automation is in fact a shadow workforce that trains the models, moderates the outputs, and handles the cases software cannot.
Why business readers should care: ghost work reframes “AI did it” into “people did it, behind an interface.” When evaluating an AI vendor or building an AI feature, the human labor in the loop - who does it, where, under what conditions and pay - is a real operational, ethical, and reputational dependency, not a rounding error.