GoKawiilStudy claims AI could cut billions of tons of CO2, but little evidence companies are doing it
A peer-reviewed study led by researchers at the London School of Economics' Grantham Research Institute and Systemiq projected AI could help reduce global emissions by 3.2 to 5.4 billion tons of CO2 annually by 2035, concentrated in the power, meat and dairy, and vehicle sectors. An AI-assisted review of public statements from major companies in those sectors found that few have announced adopting the specific emissions-cutting methods the study relied on, and almost none report achieving measurable climate results from AI so far.
GoKawiil's interpretation of the reporting above, not reported fact.
The gap between projected and actual AI adoption suggests that widely cited claims from figures like Sam Altman and Google's sustainability chief about AI's climate potential may be running well ahead of corporate practice. This could mean the 2035 emissions-reduction estimates remain largely theoretical unless companies move faster to implement AI-driven forecasting and efficiency tools in power grids, agriculture, and transport. It also raises questions about how much weight policymakers and investors should place on AI's promised climate benefits when evaluating its environmental costs.
- A 2024 study estimated AI could cut emissions by 3.2–5.4 billion tons annually by 2035, mainly in power, meat/dairy, and vehicles.
- A review of major companies in these sectors found few have adopted the specific AI methods the projections depend on.
- AI-driven forecasting in the power sector alone could theoretically prevent 1.8 billion tons of emissions a year by improving renewable energy delivery to the grid.
Source: technologyreview.com — Mary Nichols, 2026-11-09
Published there as: “The climate case for AI has not been made”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.