The AI boom has a hidden cost measured in megawatts. Training a single large language model can consume as much electricity as 100 American homes use in a year. Running inference — the process of generating AI responses for millions of users — is even more energy-intensive at scale. As AI adoption accelerates, data center energy consumption is emerging as a major environmental and regulatory issue, and the lobbying to shape energy policy has surged 233% since 2023.
The numbers are sobering. The International Energy Agency estimates that global data center electricity consumption will reach 1,000 terawatt-hours by 2026 — roughly equal to Japan's total electricity use. In the United States, data centers already consume about 4% of total electricity, and that figure is projected to reach 8-10% by 2030, driven primarily by AI workloads. Grid operators in Virginia (home to "Data Center Alley" in Northern Virginia), Texas, and Georgia have warned that data center growth is straining local power infrastructure and could lead to reliability issues.
The lobbying response has been enormous. Energy companies, data center operators, and AI firms spent a combined $67 million on energy-related lobbying in 2025 — up 233% from 2023. Much of this spending targets two goals: securing favorable electricity rates for data centers and blocking environmental regulations that would constrain energy consumption. Amazon, Google, and Microsoft — all major data center operators — have been among the top spenders. See our Follow the Money page for detailed breakdowns.
States are beginning to respond. Virginia's SB 1234 would require data centers to disclose their energy consumption and water usage annually and submit energy efficiency plans. Oregon's HB 3456 proposes a moratorium on new data center construction in areas with strained power grids until infrastructure upgrades are completed. Illinois' HB 2890 would impose a per-megawatt-hour surcharge on data centers exceeding certain energy thresholds, with revenue directed to renewable energy development. Track these bills in our state tracker.
The tech industry has pushed back with a combination of renewable energy commitments and arguments about economic benefits. Google claims to match 100% of its data center energy with renewable purchases. Microsoft has signed the largest corporate clean energy deal in history. Amazon pledges to reach net-zero carbon by 2040. But critics note that renewable energy matching is not the same as running on renewables — data centers still draw from the same grid as everyone else, and their growing demand often results in fossil fuel plants running longer. The gap between corporate sustainability marketing and actual grid impact is a growing area of scrutiny.
Nuclear energy has emerged as a surprising flashpoint. Microsoft signed a deal to restart a unit at Three Mile Island to power its AI data centers. Amazon and Google have invested in small modular nuclear reactors. These moves have drawn both praise from climate advocates who see nuclear as carbon-free baseload power and opposition from environmental groups concerned about nuclear waste and safety. The AI industry's energy demands are reshaping not just electricity markets but the entire debate about America's energy future.