A new UN study finds that Hong Kong’s data centres are among the most carbon-intensive in the world, attributing the high footprint to the city’s reliance on an electricity grid powered largely by fossil fuels. The report, titled “Environmental Cost of AI’s Energy Use,” is produced by the United Nations University Institute for Water, Environment and Health and evaluates the environmental impacts of infrastructure that powers artificial intelligence (AI). The study does not focus only on greenhouse-gas emissions. It also quantifies additional footprints linked to AI electricity generation, including effects on water use and land, alongside carbon impacts. The report compares Hong Kong’s performance with global benchmarks and says the city’s data-centre-related carbon footprint exceeds the global average. The study looks ahead to potential future impacts from continued growth in AI-related energy demand and frames its findings as a prompt for governments and industry to consider strategies that reduce “unintended impacts” associated with running AI systems on energy-intensive infrastructure.