AI’s widespread adoption is set to accelerate, driving more than $1 trillion in spending on bigger and more resource-intensive data centers in coming years. At a time of growing misinformation about AI and technology companies’ efforts to brand data centers as “green”, Bloomberg conducted two definitive investigations in 2024, using innovative data analysis and visual storytelling, which uncovered alarming consequences, hidden from the public eye, that are already affecting global energy grids and local communities worldwide:
The Energy Impacts of AI
First, reporters combined exclusive data on data centers’ locations and capacity with data on countries’ energy generation by fuel type, to investigate the popular narrative pushed by technology and AI companies that AI data centers are “green”. We quantified the energy demands of data centers globally at an unprecedented scale and detail, to uncover an alarming trend: in many countries, data center energy demand is set to exceed these countries’ entire renewable energy supply.
AI’s Hidden Impact: Distorted Electricity and Worsened Power Quality
Next, reporters analyzed exclusive and extremely granular US data on an important power quality metric called “harmonic distortion” to uncover the first-ever definitive link between data centers and distorted power in nearby homes, which can destroy electrical appliances and increase the risk of fire during a voltage surge.
Both stories made extensive use of custom data visualizations to present these important findings.
These investigations are part of an exclusive investigative series that exposes the multi-faceted impacts that AI data centers have on people, the environment, and the world around us. While published slightly outside the timeframe of this award submission, we’d like to mention the third part of this series (How AI Demand Is Draining Local Water Supplies), which uncovered yet another alarming trend: globally, the vast majority of water-thirsty AI data centers are planned or built in areas with scarce water resources and that already experience extremely high levels of water stress.
