Study Says AI Could Drive 13 Times More Fossil Fuel Emissions Than Data Centers
Updated
Updated · Futurism · Aug 14
Study Says AI Could Drive 13 Times More Fossil Fuel Emissions Than Data Centers
3 articles · Updated · Futurism · Aug 14
Summary
A new NPJ Climate Action paper says AI's climate damage extends well beyond server power use, with AI-enabled fossil fuel emissions estimated at 3 to 13 times data-center emissions.
Will and Holly Alpine found AI is helping oil and gas companies raise productivity, target drilling and expand output, with those enabled emissions exceeding AI-linked emissions cuts from renewable-energy applications.
The paper argues climate accounting has focused too narrowly on operational power demand, missing how AI reshapes fossil fuel supply economics through direct partnerships and industry-specific engineering work.
Examples cited include Microsoft's 20-year deal with Chevron for a large natural-gas plant and Amazon's buildout of what the report describes as the largest fossil fuel plant in the United States.
The authors call for tighter guardrails on AI deployment, saying opposition to new data centers should broaden into scrutiny of what AI is used for and how it pushes the economy toward a climate-negative path.