Ranked: How Dirty Are the Eight Biggest Names in AI

Big Tech giants are increasingly distancing themselves from their carbon neutrality pledges. In 2019, Amazon aimed for carbon neutrality by 2040, while Google and Microsoft committed to becoming carbon-free by 2030. However, their emissions have increased by double-digit percentages in the past six years due to the scaling of data centers for AI purposes.

To meet their growing energy demands, these companies are investing in various energy sources, some clean and others very dirty. Amazon is constructing a gas-fired power plant in Texas, which would be the country's largest source of power-related emissions if built as planned. The company remains committed to decarbonization despite this.

Among the eight leading AI firms tracked by Climate Power's Data Center Air Pollution Tracker, Amazon ranks as the cleanest, scoring 68 out of 100. The dirtiest AI firm is Elon Musk's xAI, with a score of just 6 due to its gas-powered BTM data centers, releasing a shocking 898 pounds of CO2 per megawatt-hour. Oracle follows with a score of 57, making xAI nearly ten times dirtier.

The other companies' scores range from 60 (OpenAI) to 64 (Microsoft and Google). While these scores may not represent total emissions, the three cleanest companies' projects still emit as much CO2 as a third of France annually. Critics argue that the companies' claims of sustainability are mere marketing strategies, and governments should consider these expanding carbon footprints when the same firms tout AI solutions to combat climate change.

Additionally, the sector's extreme opacity and covert actions may lead to significant underestimation of these firms' emissions.