AI could help fossil fuel companies create more emissions

New research looks at a side of AI's climate impact that gets less attention than data center power use: how AI helps oil and gas companies find and pump more fossil fuels. Oil and gas firms have used AI for decades to make extraction and refining more efficient. The study, by two former Microsoft sustainability employees, models how much that efficiency boost could raise global emissions.

The findings are striking. AI-driven productivity gains in the fossil fuel industry could push up global energy-related emissions by 1.2 to 4.8 percent a year. At the low end, that's roughly equal to Mexico's yearly emissions; at the high end, it rivals Russia's, the world's fourth-largest emitter. That's more than current projections for emissions from the global data center buildout, and it outweighs the climate benefits AI provides by improving solar, wind, and other clean energy technology.

The researchers point to a deepening relationship between tech and oil companies as evidence: Chevron is building a large gas plant in Texas to power Microsoft data centers, and Chevron's own executives have said they'll use some of that computing power to run AI inside their company. An outside energy researcher who reviewed the study called it a credible attempt to answer a question that matters for anyone tracking where climate progress actually stands.

There's no direct action for homeowners here. But it's a reminder that AI's climate footprint extends well beyond the electricity used to run data centers, into how it shapes fossil fuel supply itself.