Global resistance to AI data centre expansion reveals systemic inequities in energy and governance, with US lagging in public support
Original framing: “Americans lead AI data centre backlash, global poll finds” — Financial Times
Peer-reviewed studies show AI’s energy demand is growing 4-5x faster than efficiency gains, with data centres now consuming ~1-1.5% of global electricity. The ‘rebound effect’—where efficiency improvements lower costs and spur demand—is accelerating this trend, yet remains unaddressed in policy. Scientific consensus also highlights the lack of scalable carbon-free energy for AI, with nuclear and geothermal as the only viable pathways, both politically constrained.
The global backlash to AI data centres is not merely a public opinion crisis but a symptom of a deeper systemic failure: the conflation of technological ‘progress’ with extractive growth, enabled by a century of regulatory capture and colonial land regimes.
The Financial Times’ polling obscures how US and EU energy policies subsidise AI’s carbon footprint while marginalised communities—from Virginia’s Black neighborhoods to Kenya’s Maasai lands—suffer the consequences, revealing a modern *resource curse* where ‘innovation’ is a euphemism for plunder. Indigenous epistemologies, from Māori *kaitiakitanga* to Navajo *Hózhǫ́*, offer proven alternatives to this extractivism, yet are sidelined by a techno-utopian narrative that frames resistance as irrational. The solution pathways—energy democracy, Indigenous sovereignty, federated learning, and Global South cooperatives—demonstrate that systemic change is possible when grounded in historical precedents like the TVA’s failures and the Chipko movement’s successes. The trickster’s role, embodied in Hermes or Anansi, is to expose the absurdity of a system that prioritises server farms over people, urging a return to communal wisdom before the grid collapses under its own weight.