The rapid expansion of artificial intelligence is expected to transform the United States’ power landscape over the next decade, with data centres projected to account for nearly 20 per cent of the country’s total electricity consumption by 2035, according to a new report. The findings underscore the growing pressure that AI-driven infrastructure is likely to place on the nation’s electricity grid as technology companies continue investing heavily in advanced computing facilities.
The report suggests that electricity demand from data centres could rise at an unprecedented pace as businesses increasingly rely on AI applications requiring massive computing power. These facilities, which process and store enormous amounts of digital information, are already among the fastest-growing consumers of electricity. The surge is being fuelled by the deployment of large language models, cloud computing services and generative AI technologies that require thousands of high-performance processors operating continuously.
Experts estimate that by 2035, data centres could consume around one-fifth of all electricity generated in the United States, a dramatic increase from their current share. The projected growth is expected to outpace electricity demand from several traditional sectors, compelling utilities to rapidly expand generation capacity, upgrade transmission infrastructure and modernise ageing power grids.
The increasing appetite for electricity has prompted concerns among policymakers, energy providers and environmental groups. Industry analysts warn that meeting future demand will require substantial investments in new power plants, renewable energy projects and transmission networks. Failure to expand infrastructure in time could lead to higher electricity prices, supply constraints and increased pressure on existing grids, particularly in regions experiencing rapid data-centre development.
Technology companies have accelerated construction of hyperscale data centres across the country to support AI development. These massive facilities house thousands of specialised chips that perform complex computations around the clock. Besides consuming enormous amounts of electricity, such centres also require sophisticated cooling systems to maintain safe operating temperatures, adding further pressure on energy and water resources.
The report notes that while artificial intelligence promises significant economic benefits through improved productivity, innovation and automation, its infrastructure requirements present major challenges. Utilities are now being forced to reassess long-term electricity demand forecasts after years of relatively stable consumption. The rise of AI has emerged as one of the biggest drivers behind renewed growth in electricity use across the United States.
Industry leaders argue that the expansion of AI infrastructure can also accelerate investment in cleaner sources of electricity. Many technology firms have announced commitments to power their facilities using renewable energy, nuclear power and other low-carbon sources. However, experts caution that renewable generation alone may not be sufficient to meet the anticipated surge in demand, making a diversified energy mix essential in the coming years.
The growing concentration of large data centres has also sparked debates in several American states over electricity pricing and resource allocation. Some regulators fear that the costs of expanding the grid to accommodate AI infrastructure could ultimately be passed on to residential consumers. In response, policymakers and utilities have begun exploring mechanisms to ensure that technology companies bear a greater share of infrastructure costs while protecting household electricity bills.
Recent forecasts by the US Energy Information Administration also indicate that national electricity consumption is likely to reach record levels in 2026 and 2027, driven largely by the rapid growth of AI-powered data centres, cryptocurrency operations and increasing electrification across various sectors. These trends reinforce expectations that electricity demand will continue climbing over the next decade as digital technologies become more deeply integrated into the economy.
The report concludes that while artificial intelligence is poised to become a major engine of economic growth and technological advancement, its long-term success will depend heavily on the country’s ability to expand reliable, affordable and sustainable electricity infrastructure. Without significant investments in power generation and grid modernisation, the AI revolution could face energy constraints that may limit its future growth.
