UN Report: AI Boom Drives 150% Surge in Tech Giants’ Indirect Emissions
A new United Nations report revealed on Thursday that indirect carbon emissions from the operations of four major AI-driven tech giants—Amazon, Microsoft, Alphabet, and Meta—rose by an average of 150% between 2020 and 2023. The sharp increase is largely driven by the vast energy demands of data centers powering artificial intelligence systems.
The report, published by the International Telecommunication Union (ITU), the U.N.’s digital technologies agency, analyzed the greenhouse gas emissions of 200 leading digital companies over the three-year period. Indirect emissions include those generated from purchased electricity, heating, cooling, and steam consumed by a company’s operations.
Among the companies surveyed, Amazon posted the largest rise, with operational carbon emissions soaring 182% over the period. Microsoft followed with a 155% increase, while Meta and Alphabet saw rises of 145% and 138%, respectively.
The growing reliance on AI has led to surging energy demands, with electricity consumption from data centers growing four times faster than overall global electricity usage, according to the ITU. The report projects that carbon emissions from top-emitting AI systems could eventually reach 102.6 million tons of carbon dioxide equivalent annually, further straining existing energy infrastructures.
In response, several companies highlighted their ongoing sustainability efforts. Meta referred Reuters to its sustainability report, stating that it is taking steps to reduce emissions, energy use, and water consumption in its data centers. Amazon emphasized its investments in carbon-free energy projects, including both nuclear and renewable sources. Microsoft pointed to its recent progress in improving energy efficiency, including transitioning to chip-level liquid cooling technologies that consume less energy than traditional cooling systems.
However, the ITU noted that while more digital companies are setting ambitious emissions targets, many of these commitments have yet to translate into meaningful reductions in actual emissions. The report underscores the growing challenge of balancing AI’s rapid expansion with environmental sustainability.