Artificial intelligence is being framed less as a technological revolution and more as a national survival project. Yet even as the White House describes AI as a “new Manhattan Project,” lawmakers are still receiving primers on basic concepts.
On August 26, the Congressional Research Service issued a memo explaining to members of Congress what a data center is, how many exist, and how much electricity they consume. The fact that such a document was necessary highlights the pace at which technological expansion is outrunning institutional comprehension.
According to the memo, there are roughly 7,000 data centers worldwide, with the highest concentration in the United States, particularly in Northern Virginia and Texas. In 2022, U.S. data centers consumed about 176 terawatt-hours of electricity, accounting for nearly 4 percent of national demand. Projections suggest an additional 35 to 108 gigawatts of demand by 2030. At the midpoint estimate—50 gigawatts—that would equal the consumption of every household in California. That level of growth collides with an electric grid already under strain from plant retirements and the volatility of renewable generation.
The executive branch has been treating the issue with urgency. In January 2025, the White House declared a National Energy Emergency, followed by an April order tasking the Department of Energy (DOE) with modeling how AI demand would reshape the grid. By July, DOE warned that “retirements plus load growth increase risk of outages by 100x. Status quo is unsustainable.” Within weeks, federal lands were opened to accelerated data center permitting, and the White House released an AI Action Plan, framing the technology as a geopolitical race. Energy Secretary Chris Wright described the effort as “accelerating the next Manhattan Project—ensuring U.S. AI and energy leadership.”
For households, the impact is less abstract. In Columbus, Ohio, utility customers on standard plans saw monthly bills rise by about $20—roughly $240 annually—linked directly to data center demand. Across New Jersey, Pennsylvania, and Ohio, households faced increases of $10 to $27 per month this summer. In Oregon, utilities warned regulators that customers needed protection from rate hikes tied to new facilities. Mid-Atlantic regulators have projected 20 percent increases in household electricity costs by 2025, citing data centers as a primary driver. These local realities point to a disconnect: citizens are already feeling the financial consequences while Washington debates fundamentals.
AI’s physical footprint is often obscured by the language of the “cloud.” But each chatbot query carries real energy costs. A traditional Google search consumes about 0.3 watt-hours, while an AI query may use ten times more, around three watt-hours. Training a single large model consumes as much electricity as hundreds of American homes use in a year. Despite efficiency gains in computing, chips designed for AI workloads are drawing more power per server. Nvidia’s latest processors, for example, consume significantly more kilowatts than earlier generations. Companies have also become less transparent about energy use since ChatGPT’s launch in 2022, leaving analysts reliant on estimates rather than disclosures.
This rising demand collides with a fragile grid. The United States operates three interconnected supergrids—the Eastern, Western, and Texas systems—coordinated by regional operators such as PJM, MISO, CAISO, and ERCOT. DOE now warns that without intervention, blackout risks could rise one hundredfold by 2030. The options are politically fraught: extending the life of coal and nuclear plants, accelerating renewable and transmission build-out far beyond historic levels, or slowing data center construction to areas with abundant supply. Each choice carries costs, but DOE’s assessment leaves little ambiguity: the current trajectory points to higher rates and greater instability.
The broader context is a debate within the scientific community itself. Figures such as Geoffrey Hinton, who left Google in 2023, warn of AI as an existential risk to humanity. Others counter that AI could accelerate medical breakthroughs or even help mitigate climate change. The disagreement is not marginal but fundamental, and it complicates policymaking. Unlike the original Manhattan Project, which was secretive and narrow in scope, AI is unfolding publicly across sectors, reshaping jobs, media, commerce, and household expenses simultaneously. In 2025 alone, Google, Meta, Amazon, and Microsoft are projected to spend more than $350 billion on AI data centers—nearly double the inflation-adjusted cost of the Apollo program.
Stay updated on the latest in energy! Follow us on LinkedIn, Facebook, and X for real-time news and insights. Don’t miss out on exclusive interviews and webinars—subscribe to our YouTube channel today! Join our community and be part of the conversation shaping the future of energy.
