AI Could Worsen Climate Crisis by Boosting Fossil Fuel Production
Research reveals AI use in oil and gas industry may significantly increase global emissions, outpacing its benefits for clean tech.
1The Dark Side of AI
New research by former Microsoft sustainability workers shows AI's efficiency gains in fossil fuel industry could exacerbate climate crisis. Published in npj Climate Action, the study indicates AI-assisted oil and gas production could increase annual emissions equivalent to Mexico's total output. In worst-case scenario, this rise could match Russia's emissions - the world's fourth largest polluter.
2Counterproductive Effect
The research reveals AI's contributions to clean energy technologies pale in comparison to emission increases it enables in fossil fuel sector. Surpassing even data center energy consumption projections, this growth threatens climate targets. Researchers Will and Holly Alpine, who left Microsoft, emphasize tech companies focus on operational emissions while ignoring 'enabled emissions'.
3AI in Industry
Oil and gas companies have used various AI forms for decades to locate and extract underground resources more efficiently. New economic modeling shows AI acts as productivity booster across all fossil fuel sectors from extraction to refining to power generation. Companies like Chevron are both powering data centers and enhancing their own AI capabilities through partnerships with Microsoft.
4Macroeconomic Modeling
Researchers developed complex economic model based on efficiency gain reports from fossil fuel companies using AI tools. The model predicts AI could increase global energy-related emissions by 1.2% to 4.8% - exceeding many countries' annual emission targets. Energy researcher Jon Koomey notes the conclusions appear robust.
5Chevron-Microsoft Partnership
Chevron's planned large-scale natural gas plant in Texas to power Microsoft data centers highlights symbiotic relationship between AI and fossil fuel industry. Chevron New Energies president Jeff Gustavson revealed the collaboration will also strengthen company's AI capacity, with portion of plant's energy dedicated to Chevron's AI operations.
6Unmeasured Impact
While tech companies measure supply chain emissions, they don't track how their tools increase fossil fuel production. The Alpines term this 'enabled emissions'. They argue Microsoft's sustainability measures focus largely on operational emissions while enabled emissions cause far greater climate damage.
7Double-Edged Sword
Energy researcher Jon Koomey notes machine learning can make data center cooling 30-40% more efficient while simultaneously making fossil fuel extraction much cheaper and faster. He states net effect remains unknown but praises new research as credible macroeconomic modeling attempt to answer this critical question.
8Climate Goals at Risk
Findings suggest widespread AI use in fossil fuel industry could jeopardize 1.5°C global warming limit. This 'self-reinforcing cycle' between tech and energy giants slows clean energy transition. Experts warn AI developers must consider impacts across all industries, not just renewables and efficiency.
Original article: Wired — Yapay Zeka Fosil Yakıt Üretimini Artırarak İklim Krizini Kötüleştirebilir