Artificial intelligence could increase carbon emissions by nearly 5 percent by making fossil fuel production more efficient, according to new research. The finding shifts attention from power-hungry data centers to AI’s wider use across the energy industry.
The study suggests that AI may help oil, gas, and coal companies find resources, improve extraction, reduce operating costs, and increase output. Those gains could encourage greater fossil fuel use, producing far more emissions than AI data centers generate directly.
Productivity Gains Carry Climate Costs
AI systems can process large volumes of geological, equipment, and market data. Fossil fuel producers may use those tools to identify deposits, predict maintenance needs, manage pipelines, and refine drilling plans.
Each application can make production cheaper or more reliable. However, higher productivity may also extend the life of fossil fuel projects that would otherwise become too costly.
By making the fossil fuel industry more productive, AI could help increase carbon emissions by up to nearly 5 percent.
The estimate represents a potential increase, rather than a guaranteed outcome. Actual emissions would depend on how widely companies adopt AI, how energy markets respond, and whether governments tighten climate policies.
The limited research summary does not identify the period covered, the baseline emissions level, or the assumptions used to calculate the upper estimate. Those details would be needed to judge how likely the 5 percent scenario may be.
Data Centers Are Only Part of the Impact
Public debate about AI’s climate effects has focused mainly on data centers. Training and running large AI models requires electricity, while cooling systems can consume water and additional power.
The new finding points to a larger concern: AI’s indirect emissions may exceed its operational footprint. A system can use electricity in a data center while also helping another industry produce and sell more carbon-intensive energy.
That distinction matters for climate accounting. Measuring only electricity use may miss emissions enabled by AI-supported business activity.
- Direct effects include electricity and water used by computing facilities.
- Indirect effects include higher fossil fuel output linked to improved productivity.
- Policy effects depend on regulations, carbon prices, and clean-energy investment.
Efficiency Does Not Always Reduce Demand
Efficiency is often presented as an environmental benefit because it reduces the resources needed for each unit of production. Yet lower costs can also increase total demand and output.
This pattern is sometimes called the rebound effect. If AI reduces the cost of extracting oil or gas, producers may develop reserves that were previously uneconomic. Buyers may also consume more if prices fall.
The same technology could support climate goals in other settings. AI can help balance electric grids, forecast renewable generation, detect methane leaks, and improve energy use in buildings. These benefits do not erase added fossil fuel emissions, but they show that the outcome depends on where and how AI is deployed.
Policymakers Face a Wider Accounting Challenge
The research suggests that governments and companies may need to assess AI projects through their full economic effects. Data-center efficiency standards alone would not address increased extraction or consumption enabled by AI.
Possible responses include requiring companies to report AI-related energy impacts, strengthening methane controls, and testing whether productivity tools conflict with emissions targets. Investors may also seek clearer disclosure about how AI spending affects fossil fuel output.
The central finding is that AI’s climate impact cannot be measured solely through servers and electricity bills. If productivity gains increase fossil fuel supply, indirect emissions could become the larger risk. Future research will need to test the nearly 5 percent estimate and identify which uses of AI produce the greatest effects.
Deanna Ritchie is a managing editor at DevX. She has a degree in English Literature. She has written 2000+ articles on getting out of debt and mastering your finances. She has edited over 60,000 articles in her life. She has a passion for helping writers inspire others through their words. Deanna has also been an editor at Entrepreneur Magazine and ReadWrite.
