US data centers could use more natural gas than Germany and Japan combined by 2035

TechCrunch reported on September 15 that the AI boom could push US data centers to consume more natural gas than Germany and Japan combined by 2035.

According to the report, sustained growth in AI computing could make American data centers one of the largest consumers of natural gas in the world. The claim connects AI infrastructure directly to energy supply: the more compute training and inference demand, the more electricity and fuel sit behind it. Data centers are no longer only a chips-and-software story.

Natural gas keeps entering the conversation because gas-fired plants can be built faster than many alternatives and can ramp output up and down with demand, which makes them a common answer to a fast-growing power gap. The report does not detail which projects drive the increase, but it points to a broader shift. The bottleneck for AI infrastructure is moving from chips to energy, and energy problems are often harder to solve quickly than chip problems.

For the United States, that means data center demand could compete with household heating, industry and existing power generation for the same fuel. Natural gas is both a power source and a heating fuel, so once data centers become a major buyer, questions about prices, pipeline capacity and who absorbs regional power costs move to the front.

The comparison in the headline says something on its own: Germany and Japan are both major natural gas markets, and measuring US data centers against their combined demand suggests the projected volumes reach well beyond ordinary sector growth. It is also why data center energy use has moved from a technical detail to a public debate.

The caveat is that this is a projection for 2035, not a recorded outcome. Construction timelines, chip efficiency, model efficiency and shifts in the power mix will all shape the final number, and the width of such forecasts reflects that uncertainty.

Three things are worth watching: how the added gas demand shows up in concrete projects, whether it puts pressure on gas and electricity prices, and whether data centers can raise the share of renewables and storage fast enough to keep AI growth from being shifted onto the energy system.