As artificial intelligence reshapes industries and daily life, concerns about its environmental impact are mounting. The AI revolution’s voracious appetite for energy presents an increasingly urgent challenge for climate goals and sustainability efforts.

The Escalating Energy Demand

AI technologies are driving an unprecedented surge in energy consumption. Data centers, once relatively modest consumers of electricity, have seen their footprint expand dramatically since the AI boom began.

“The data center demand used to be about one to two percent of the United States’ electricity consumption,” explains Vijay Gadepally, senior scientist at the MIT Lincoln Laboratory Supercomputing Center. “As of today, we’re somewhere around four or five percent.”

This rapid escalation marks a significant shift for the United States, where energy use had remained relatively flat for two decades prior. According to a recent International Monetary Fund report, this trend necessitates careful alignment of energy policies with AI development to support technological innovation while mitigating environmental consequences.

Quantifying the Environmental Impact

The scale of the environmental challenge is becoming clearer. According to The Register, datacenters are predicted to emit 2.5 billion tons of greenhouse gases by the end of this decade—triple their current output. This dramatic increase can be attributed largely to the proliferation of energy-intensive AI applications.

Food & Water Watch’s recent analysis paints an even more concerning picture of the near future. Their research indicates that energy demand from AI servers and data centers in the U.S. alone is expected to increase up to threefold between 2023 and 2028. By 2028, American AI infrastructure could consume approximately 300 terawatt-hours (TWh) of energy annually—equivalent to powering over 28 million households.

Beyond Electricity: The Water Footprint

The environmental concerns extend beyond electricity consumption. MIT News highlighted that generative AI’s environmental impact includes substantial water requirements for cooling high-performance computing systems.

Food & Water Watch quantifies this water demand, projecting that by 2028, AI could require as much as 720 billion gallons of water annually just for cooling servers—equivalent to over one million Olympic-size swimming pools, or enough to meet the indoor water needs of 18.5 million American households.

This water consumption is particularly problematic given that many AI data centers are located in regions already experiencing water scarcity, creating potential conflicts with other essential water uses.

Climate Implications

MIT’s Climate Portal has characterized AI as creating “an unexpected surge in energy demand, and with it, climate-warming greenhouse gas emissions.” The portal notes that addressing this challenge will require a three-pronged approach: developing more computationally efficient AI models, designing more energy-efficient data centers, and generating new clean energy sources to power it all.

For context, computing had historically been a relatively minor climate polluter, with data centers and transmission networks producing just 0.6% of global greenhouse gas emissions as recently as 2020. The rapid AI-driven expansion threatens to significantly increase this contribution.

Industry Response and Path Forward

As awareness of these environmental challenges grows, pressure is mounting on the tech industry to address the sustainability of AI development. The solution likely involves a combination of technological innovation, infrastructure improvements, and policy adjustments.

The coming years will be critical in determining whether the AI revolution can be reconciled with global climate goals, or whether unchecked growth in energy-intensive computing will undermine progress toward a sustainable future.