Official data already shows that US data centers consumed about 176 terawatt-hours of electricity in 2023, representing roughly 4.4% of total electricity demand, and that figure is expected to surge dramatically as AI and cloud infrastructure expand. This is not a linear trend. Between 2014 and 2023, power consumption from data centers more than tripled, and projections suggest usage could reach between 6.7% and 12% of total U.S. electricity consumption by 2028.
So when analysts warn that data centers could soon approach 7% or more of total electricity demand, they are not exaggerating. Some forecasts from the Electric Power Research Institute even estimate that data centers could consume up to 9% of U.S. electricity generation by 2030.
The key issue is that this surge is being driven largely by AI infrastructure and digital services, which are far more energy intensive than traditional computing. Advanced AI servers alone require substantially more power, often using two to four times the watts of conventional hardware, while massive campuses under construction can consume gigawatts of power, which is equivalent to millions of homes.
The industrial revolution demanded coal, the 20th century demanded oil, and now the digital age demands electricity. The US Energy Information Administration has already warned that electricity demand growth is accelerating after decades of stagnation, driven significantly by the expansion of data centers. Electricity infrastructure in the United States was not designed for computational loads operating 24/7. Utilities are now being forced to expand grid capacity, invest in transmission, and upgrade generation just to accommodate data center demand, which in many regions is becoming the dominant driver of new electricity consumption.
This will inevitably translate into higher electricity costs. Studies already show that grid upgrades and capacity pressures linked to data centers can push up residential electricity bills and strain regional power markets. That means households will ultimately subsidize the digital infrastructure of Big Tech through rising utility costs. This is a hidden tax most people do not yet see.
Capital is flowing into AI, cloud computing, and digital infrastructure at an unprecedented pace. But capital concentration into one sector always creates bottlenecks elsewhere. In this case, the bottleneck is energy. This trend confirms a broader shift in the global economy. We are replacing physical production with digital computation, yet computation is not energy free. In fact, it is becoming one of the most energy-intensive sectors of the modern economy.
This is why the energy crisis of the future may not begin with oil shortages, but with electricity shortages. The digital economy is energy dependent, and as confidence shifts into AI and computational systems, the real foundation of economic power becomes the electrical grid itself. Those who control reliable energy will ultimately control the next phase of economic growth.