AI Data Centers Could Consume 20% of U.S. Electricity by 2035

The rapid expansion of artificial intelligence (AI) is driving an unprecedented surge in electricity demand from data centers worldwide. According to a recent report by market research firm BloombergNEF, data centers alone could consume 20% of all electricity generated in the United States by 2035—nearly four times their current power usage.

AI Boom Driving Massive Energy Demand

The report estimates that global data center power capacity will reach nearly 200 gigawatts over the next decade as AI capabilities continue to expand. Roughly half of that capacity will be used for AI model training and inference. By 2033, the United States is expected to account for approximately 64% of the world's AI chip deployment, making it the largest consumer of AI computing power.

Growing Pressure on the National Power Grid

The soaring electricity demand is already placing significant strain on the U.S. power infrastructure. Data centers are projected to consume 34% of the electricity supplied by the PJM Interconnection—which serves states from Virginia to Illinois—and 22% of the ERCOT grid in Texas. The widening gap between electricity supply and demand has already contributed to power prices rising by as much as 76% in several regions over the past year.

Global Impact Extends Beyond the U.S.

The surge in AI-driven electricity demand is not limited to the United States. If current adoption trends continue, data centers worldwide are expected to require an additional 1,935 terawatt-hours (TWh) of electricity annually by 2033. That figure is roughly equivalent to India's total annual electricity consumption.

Compared with earlier forecasts, the projected growth in power demand has accelerated significantly. As AI infrastructure continues to expand, balancing technological progress with environmental sustainability while keeping electricity affordable for consumers is becoming an increasingly urgent global challenge.

//DBTech/BMT/OR//