The Dark Side of AI: Unveiling the Climate Impact
In a world where artificial intelligence (AI) is often hailed as a revolutionary force, a recent study sheds light on its unintended consequences. The focus? The oil and gas industry's use of AI to locate and exploit new energy reserves, a practice that significantly worsens climate pollution.
The Study's Findings
Will and Holly Alpine, former Microsoft employees turned activists, have published a groundbreaking study in Nature. Their research reveals that the "enabled emissions" resulting from AI-assisted oil and gas drilling are far more detrimental to the climate than the emissions from the data centers powering AI platforms. This finding challenges the common perception of AI as a climate-friendly technology.
The study estimates that AI tools have increased global emissions by a staggering 0.47 to 1.8 gigatonnes of carbon dioxide annually, which is up to 13 times the emissions attributed to data centers. This boost in emissions is due to AI's ability to make oil and gas production faster, cheaper, and more profitable, leading to expanded output that would otherwise be unfeasible.
Economic Simulation and Productivity Shocks
To reach these conclusions, the Alpines employed an economic simulation model. They analyzed the effects of AI on both the fossil fuel and renewable energy industries, converting these impacts into "productivity shocks." By considering various levels of AI adoption and aligning their findings with respected forecasts, such as the International Energy Agency's 2035 projections, they were able to quantify AI's asymmetric impact on the fossil fuel industry.
AI's Dual Role
Will Alpine, the study's lead author, acknowledges that AI can indeed advance renewable energy, strengthen grids, and improve efficiency. However, he emphasizes that AI has been boosting the productivity of the fossil fuel industry for years, acting as an economic lever that reinforces the dominance of fossil fuels.
Holly Alpine adds that most measures of AI's climate impact are framed as a trade-off between data center energy use and the potential emissions reductions AI might offer. She argues that this perspective overlooks how AI makes oil and gas production more commercially viable, an aspect that needs recognition and governance to address AI's full climate impact.
The Tech-Fossil Fuel Nexus
The Alpines' study exposes a troubling alliance between the tech and fossil fuel industries. AI developers are actively making deals with oil companies, selling proprietary tools designed to accelerate oil production. As Clara Vondrich from Public Citizen puts it, "Big Tech companies, self-avowed climate champs, are working hand-in-glove with Big Oil to find, dig, and burn more fossil fuels."
AI's Potential for Good
It's important to note that AI is not inherently evil. Another study published in the same journal in 2025 found that AI has the potential to reduce emissions in the food, power, and mobility industries, which collectively account for nearly half of global greenhouse gas emissions. AI can speed up the transition away from fossil fuels and reimagine interconnected systems. However, there is a lack of research on the combined effects of AI and the low-carbon transition, highlighting a critical knowledge gap.
Industry Response
Microsoft, a tech giant, has acknowledged the evolving context of AI and its commitment to becoming a carbon-negative, water-positive, and zero-waste company that protects ecosystems. The company believes that technology plays a critical role in enabling industry-wide decarbonization, but this progress must be balanced with today's energy needs and tomorrow's innovation.
Conclusion
The Alpines' study serves as a wake-up call, highlighting the urgent need to address AI's enabled emissions. As Holly Alpine states, "Until enabled emissions are recognized, measured, and governed, we’re only addressing a fraction of AI’s climate impact." It's a call to action for both the tech and energy industries to prioritize sustainability and ensure that AI's potential for good is realized without exacerbating the climate crisis.
