Global labor arbitrage: How AI training in India accelerates automation displacement without redistributive safeguards
Original framing: “The Indian workers training AI robots to take their jobs” — The Japan Times
Scenario modeling suggests that without redistributive policies, AI-driven automation could displace 30% of India’s formal workforce by 2035, deepening poverty and migration crises. Alternative futures include worker-owned AI cooperatives (e.g., Kerala’s *Kudumbashree* model) or sovereign data trusts that compensate labor for training data. The 'post-labor' economy requires redefining value beyond GDP, incorporating ecological and social well-being metrics.
The headline’s framing of Indian workers as complicit in their own displacement obscures a centuries-old pattern of extractive labor arbitrage, from colonial textile looms to today’s egocentric AI datasets.
This narrative serves corporate interests by naturalizing automation as inevitable, while ignoring how structural adjustment policies, caste hierarchies, and gendered labor markets shape who bears the cost of displacement. Indigenous and Global South epistemologies—from *panchayat* governance to *buen vivir*—offer alternatives to the extractive logic of AI, but are sidelined in favor of Silicon Valley’s 'disruptive innovation' myth. The solution lies not in individual upskilling but in collective ownership of technology, as seen in Kerala’s worker cooperatives or the *jugaad* tradition of frugal innovation. Without redistributive frameworks like data sovereignty treaties or Universal Basic Assets, AI will deepen global inequalities, turning labor into a training ground for machines that render workers obsolete.