The Hidden Climate Cost of AI: An Eye-Opening Study on Emissions
Recent research has unveiled a troubling dynamic in the relationship between artificial intelligence (AI) and fossil fuel emissions, suggesting that AI-driven productivity gains lead to greater planet-heating pollution than the reductions offered by renewable energy improvements. This groundbreaking study analyzed the climate impact of AI across the entire power sector, showcasing the complexities often overlooked in discussions about technology’s role in energy.
AI’s Influence on the Energy Sector
Researchers modeled the potential for AI to enhance clean power generation alongside its capabilities in traditional fossil fuel extraction. Over 64 different scenarios, they discovered that net yearly carbon emissions increased by 0.47 to 1.8 gigatonnes. To put this in perspective, that represents about 1-5% of the total annual emissions produced by the energy sector.
This comprehensive approach highlights a significant gap in previous studies, which typically focused on the positive impacts of AI—like reducing downtime for renewable energy sources or optimizing electricity grids—without adequately addressing the emissions that can result from increased fossil fuel productivity.
A Shift in Perspective
Lynn Kaack, an assistant professor of computer science and public policy at the Hertie School, pointed out a crucial oversight in prior examinations of AI’s climate effects. “What most studies have done so far is compare the data center energy use with the emissions savings AI has,” she remarked. “They completely omit this picture of AI causing increases in emissions.” This shift in focus is vital for developing a more nuanced understanding of how technology interacts with climate change.
The researchers found that net emissions only decreased when AI was implemented without enhancing productivity in the fossil fuel sector. They further established that for the emissions reductions from renewable sources to outweigh those from fossil fuels, productivity gains in renewables would need to surpass those from fossil fuels by at least a factor of four.
More Read
The Rate of AI Adoption
Co-author Holly Alpine, a former Microsoft employee who has since co-founded the Enabled Emissions campaign group, noted the more conservative assumptions made during their research. They posited that AI adoption would occur at the same rate for both fossil fuel and renewable energy sectors. Yet, the reality is that fossil fuel applications have already been widely implemented, with concrete contracts and deployments already demonstrated. In contrast, many renewable applications remain in pilot phases or are limited by regulatory and logistical barriers.
Future of Fossil Fuels and AI
The International Energy Agency (IEA) estimates that AI could boost recoverable oil and gas reserves by 5% while simultaneously cutting deepwater offshore project costs by 10%. This potential has not gone unnoticed by fossil fuel executives, who are already heralding AI advancements as the next major breakthrough akin to the fracking boom.
Saudi Aramco, one of the largest oil companies in the world, revealed that AI has been integrated across its operations, leading to increased productivity and an uptick in the number of oil wells. Additionally, Equinor, another major player, attributed several oil discoveries to AI-enhanced seismic technologies, reporting their largest discovery in 2025 thanks to automated data interpretation.
Economic Valuation of AI in Fossil Fuel Exploration
Rystad Energy, a prominent research and intelligence firm based in Oslo, projected that digital technology and AI could collectively generate nearly $500 billion in value for companies focused on fossil fuel exploration and production between 2026 and 2030. This projection stems from enhanced efficiency, increased production rates, and shorter development timelines. The financial benefits seen across the industry are already evident, with operators reporting significant savings related to AI advancements.
The Impact of Data Centers on Emissions
While this study didn’t include the energy demands of data centers in its calculations, the significance of this factor cannot be overlooked. Climate scientists and energy experts have expressed concerns about the growing number of AI-powered data centers, many of which rely on fossil fuels to meet their energy needs. Researchers found that AI-driven productivity improvements in fossil fuel extraction could lead to emissions at least three times greater than what current estimates for data centers suggest.
The Broader Implications for Climate Action
Ketan Joshi, an independent climate analyst, cautioned about the broader implications of relying on AI technology in our fight against climate change, stating, “The AI sector is fundamentally hungry for fossil fuels.” This underscores a critical paradox: as we innovate in technology, we must remain vigilant against an accompanying rise in fossil fuel consumption. Joshi emphasized that simply encouraging companies to invest in renewable projects isn’t sufficient; a structural approach is necessary for meaningful climate safety.
The evolving conversation around AI’s role in the energy sector is crucial for shaping effective climate policies. The findings from this study compel us to rethink how technologies intended to improve efficiency can also exacerbate existing issues, ultimately highlighting the intricate balance we must achieve between innovation and sustainability.
Inspired by: Source
