Artificial intelligence has become the new engine of the global economy — transforming how we work, communicate, and innovate. But as the race to build smarter, faster, more capable AI systems accelerates, a pressing question emerges: how much energy does it take to power intelligence itself? And as the world builds new data centers at record speed, who will ultimately pay the price?
The Energy Behind Artificial Intelligence
For years, computing was considered a relatively small climate offender. According to MIT Climate (“AI’s Energy Use: A Big Problem for Climate Change?”), as recently as 2020, data centers and data transmission networks accounted for just 0.6% of global greenhouse gas emissions. That footprint is now growing rapidly, thanks to the explosive demand for generative AI, which requires immense computing power and round-the-clock electricity.
Vijay Gadepally, a senior scientist at the MIT Lincoln Laboratory Supercomputing Center, explained that U.S. data center demand has jumped from 1–2% of national electricity use to around 4–5% today. The U.S. Department of Energy predicts that this usage could double or even triple by 2028, with similar trajectories in China and Europe.
This sudden surge reverses decades of energy stability. For nearly twenty years, U.S. electricity use was flat, allowing the country to slowly replace coal with cleaner energy sources. But AI is now reshaping that balance, reigniting fossil fuel dependency and raising fresh concerns about emissions, water use, and grid reliability.
Source: MIT Climate – AI’s Energy Use: A Big Problem for Climate Change?
AI’s Climate Footprint and the Global Power Strain
A recent AP News report revealed that electricity consumption from data centers is expected to more than double globally by 2030, according to the International Energy Agency. The U.S. leads this surge, followed by China. In many regions, that electricity still comes from coal, oil, and natural gas — energy sources that release carbon dioxide and warm the planet.
Public concern is growing fast. The same AP-NORC poll found that Americans are more worried about AI’s environmental impact than that of meat production, aviation, or cryptocurrency mining. The reason is simple: AI requires massive, always-on facilities that devour electricity and water — and those costs ripple through entire communities.
Some tech companies have turned to nuclear energy as a “cleaner” alternative, while others quietly roll back climate commitments to meet the demands of the AI boom. Meanwhile, the U.S. government has encouraged AI development as part of a national economic and security strategy, even if it means relaxing environmental protections to speed up data center construction.
The facts are clear: artificial intelligence is fueling climate change.
Source: AP News – Americans Worry AI Will Hurt the Environment
The AI Arms Race and Its Local Consequences
As NPR’s “It’s Been a Minute” uncovered in a recent interview with Washington Post energy reporter Evan Halper, the global AI boom is being driven by a geopolitical race for dominance. The U.S., China, and Europe are each vying for technological control — but the infrastructure to support it is immense.
AI data centers are sprawling across the U.S., from Virginia to Texas to Ohio, often built on the sites of retired coal plants. Some of those coal facilities, once slated for closure, are now staying open longer or being replaced by natural gas plants to feed new AI installations. In one Pennsylvania project, seven new gas plants are being built where a coal plant once stood.
The impact is deeply local. In places like Omaha, Nebraska, communities living near coal plants have seen them remain operational years longer than promised — often in low-income or minority neighborhoods already burdened by pollution. In other regions, like Arizona and Texas, data centers are straining water resources, leading to new environmental and public health concerns.
Perhaps most worrying: as energy use spikes, residents’ power bills are rising too. In areas hosting large data centers, local electricity rates have jumped by as much as 20%, NPR reports. Utility upgrades, once the public’s responsibility, are increasingly linked to corporate energy needs. The question remains — should local families and small businesses be subsidizing AI’s power surge?
Source: NPR Transcript – AI Data Centers, Energy, and the Economy
The Economic Ripple Effect
AI is reshaping not just the digital world but the real economy. As data center energy demand grows, electricity costs, infrastructure investments, and carbon emissions are all climbing. Nations that can afford clean energy investments may weather the transition, but developing economies could face widening inequalities as energy-intensive AI industries cluster in richer regions.
And while tech leaders argue that AI will ultimately help combat climate change — optimizing energy grids, forecasting weather, and managing renewable power — the current reality paints a more complicated picture. The technology promising to save the planet may first make it harder to cool.
The Bottom Line: Who Pays for Intelligence?
Artificial intelligence is not just transforming society; it’s transforming the climate economy itself. As AI systems grow larger, the costs — both financial and environmental — are being distributed unevenly.
The next stage of the AI revolution may not be defined by algorithms or ethics, but by energy access and economic equity. The question is no longer whether AI will change the world. It already has. The real question is: who will pay the bill?
Interested more about how AI is interconnected with the rest of the world? Check out the tech section of Interconnected Earth here.