Artificial Intelligence Could Soon Outpace Bitcoin in Energy Use
The rapid rise of artificial intelligence is reshaping not just technology, but the world’s energy landscape—and the numbers are staggering. According to recent analyses, AI’s electricity demands are on track to overtake even Bitcoin mining by the end of 2025, marking a dramatic shift in how digital industries impact global power grids.
AI’s Growing Energy Appetite
AI already accounts for up to 20% of the electricity used by data centers worldwide, and this figure is expected to climb sharply. By the end of 2025, AI could consume as much power as all the data centers in the United Kingdom—roughly 23 gigawatts, or the equivalent of powering tens of millions of homes. For perspective, that’s nearly double the current electricity consumption of the entire Bitcoin mining network, which stands at about 10 gigawatts.
Data centers, the massive warehouses of servers that power everything from cloud computing to AI training, are at the heart of this trend. In 2023, these facilities consumed as much electricity as the entire country of Germany or France—up to 500 terawatt-hours globally. By 2030, that number could triple, reaching levels comparable to India’s total electricity use. In the United States alone, data center electricity demand is projected to more than triple, potentially surpassing 600 terawatt-hours by 2030.
Why Is AI So Power-Hungry?
Unlike Bitcoin mining, which relies on specialized hardware running continuously across the globe, AI’s energy needs are driven by the explosive growth in training and running large language models. Each new generation of AI chips, like Nvidia’s Grace Blackwell, is more efficient—but as models grow larger and more complex, the overall energy demand keeps rising. This phenomenon, known as Jevons’ Paradox, means efficiency gains are offset by increased usage, leading to higher total consumption.
The United States and China are leading the charge, accounting for nearly 80% of projected global growth in data center electricity use. The surge is already forcing utilities to plan new gas-fired power plants and even revive nuclear projects to keep up with demand. In the US, data centers could soon use 7.5% of the nation’s electricity, enough to power nearly a third of all American homes.
While tech companies are investing in renewable energy, much of the short-term growth will be met by fossil fuels. However, experts predict that by 2030, renewables could supply half of the energy for data centers, up from about 27% today. Europe is expected to get 85% of its new energy from renewables, but the global picture remains mixed.
Artificial intelligence is driving an unprecedented surge in electricity demand, with implications for everything from energy policy to climate change. As AI’s footprint grows, so does the urgency to find sustainable solutions—before the world’s power grids are overwhelmed by the digital revolution497.
References:
- https://www.iea.org/reports/energy-and-ai
- https://www.imf.org/en/Blogs/Articles/2025/05/13/ai-needs-more-abundant-power-supplies-to-keep-driving-economic-growth