Exploited data workers in the Global South are AI's secret ingredient
| Source: Mastodon | Original article
Over 160 million exploited data workers in the Global South form the hidden backbone of AI development.
An interview published this week shines a stark light on the human engine behind today’s artificial‑intelligence boom. Sociologist Antonio Casilli, author of *Waiting for Robots*, estimates that more than 160 million workers in the Global South are employed to label, curate and otherwise “train” data that powers everything from chatbots to autonomous‑driving systems. He argues these workers are deliberately kept out of public view, describing them as “the secret ingredient of AI itself.”
The revelation matters because it undercuts the narrative of fully autonomous, self‑sustaining algorithms. While AI products are marketed as products of pure code, their performance still hinges on massive, low‑paid human labour that is often outsourced to regions with weak labour protections. The hidden workforce raises questions about the ethical foundations of AI development, the fairness of profit distribution, and the potential for regulatory scrutiny. If the scale of exploitation is as large as Casilli suggests, it could become a focal point for activists, policymakers and investors demanding greater transparency in AI supply chains.
What to watch next are the reactions from the tech firms that rely on this hidden labour. Industry bodies may be pressured to disclose data‑annotation contracts, and legislators in Europe and North America could consider new reporting requirements for AI training pipelines. Labor organisations in the Global South are likely to amplify calls for fair wages and safe working conditions. As the debate moves from academic circles to boardrooms, the visibility of these “data workers” could reshape how the AI sector accounts for its human cost.
Sources
Back to AIPULSEN