Their conclusion, covered by The Guardian, is blunt: the technology's contribution to fossil fuel expansion currently exceeds any climate gains it produces. That's an uncomfortable finding for an industry that has spent years positioning AI as a tool for decarbonization.
Two Different AI Stories
On one side, there's the AI that gets talked about at climate conferences. Machine learning models that optimize solar and wind grids. Systems that model climate risk. Tools that improve battery chemistry research. On the other side, there's the AI that oil and gas companies are quietly scaling up. Seismic analysis. Reservoir modeling. Predictive maintenance for drilling equipment. That second category is bigger, and growing faster, according to the research. Energy companies including major oil producers have invested heavily in AI to squeeze more output from existing wells and identify new drilling sites more efficiently. The technology shortens exploration timelines and cuts costs, which makes fossil fuel extraction more profitable, not less.
The Math Problem
The core issue is scale. Renewable energy optimization and climate modeling remain relatively niche applications. Fossil fuel companies, by contrast, have deep pockets and clear commercial incentives to use AI for extraction. That imbalance means the net effect, at least for now, tilts toward more emissions rather than fewer. There's also a second, less discussed cost: AI itself is energy hungry. Data centers powering large language models and other AI systems are driving new electricity demand across the globe. In several regions, that demand has been met partly through natural gas plants, some newly built or newly extended specifically to serve AI infrastructure. So AI isn't just helping extract fossil fuels. In some cases, it's directly consuming the electricity generated from them.
A Familiar Corporate Contradiction
Tech companies have made ambitious net-zero pledges while simultaneously signing partnerships with energy firms to power AI expansion. Several major cloud providers have also sold AI tools directly to oil and gas clients, a business line that generates revenue even as those same companies publicize climate commitments elsewhere. That contradiction isn't new, but the study adds hard evidence to what critics have long suspected: sustainability marketing and actual deployment patterns don't match. None of this means AI has no climate value. Grid optimization, methane leak detection and materials science research all show genuine promise. The problem is proportion, not existence.
What This Means for Miami
Florida's economy sits at the center of climate exposure, from insurance markets to coastal infrastructure, making this tension directly relevant to South Florida business leaders. Miami's growing AI and climate-tech startup scene, along with university researchers working on sustainability and energy systems, may find opportunity in building tools that genuinely reduce emissions rather than accelerate extraction. Investors backing climate-focused AI ventures in the region should note that differentiation matters. As scrutiny of "green AI" claims increases, startups with measurable, verifiable climate impact could stand out from a market where the technology's dominant use, at least statistically, still leans fossil fuel positive.
