AI’s carbon footprint is even worse than it seems.
It’s not just that new data centers are leading to new fossil-fueled power plants—like an Amazon-backed gas plant in Texas that could be the largest single source of CO2 pollution in the U.S. AI also is making the fossil fuel industry more productive, and a new study finds that those productivity gains could generate as much as 13 times more emissions than data centers do today.
The study, published in the Nature Portfolio journal npj Climate Action, focuses on “enabled emissions”: emissions created when oil and gas companies use AI to extract and refine fossil fuels more quickly and cheaply, and to generate more power.
It also calculated how AI is making renewable energy more productive. But the effect on fossil fuels far outweighs the benefit for renewables.
For global emissions to stay roughly the same, renewable productivity would have to improve around 4-5 times more than fossil productivity. That’s partly because even a small improvement in extracting fossil fuels can lead to much more fossil fuel use.
The study found that productivity gains from AI increase global emissions by 0.47 to 1.8 gigatonnes of CO2 annually. (For reference, 1.8 gigatonnes is about three times Germany’s annual emissions.) The emissions from productivity gains for fossil fuels are between 3.3 and 13.3 times greater than the emissions from data centers today.
Two of the study authors are former Microsoft employees who left to start a nonprofit called the Enabled Emissions Campaign.
“I spent years building AI platform tools and have seen firsthand how they’re used,” Will Alpine, the lead author, said in a statement. “Like any tool, AI can accelerate whatever it’s applied to. Yes, it can advance renewable energy, strengthen the grid, and improve efficiency. But it has also been boosting the productivity of the fossil fuel industry for years, and our research shows that effect is asymmetric: it acts as an economic lever that reinforces the viability and dominance of fossil fuels.”