VoltScope 電馳誌
Energy
Charging
Smart Mobility

AI Data Centers Drive Nearly 40% of Costs in Largest U.S. Grid as Trump Administration Shifts Infrastructure Burden

The latest capacity market auction for PJM Interconnection, the largest U.S. regional grid, reveals that AI data center electricity demand accounts for 38% of total costs, amounting to $6.3 billion. The Trump administration has expanded the Ratepayer Protection Pledge, requiring large data centers to bear the costs of new power infrastructure, a move that has drawn skepticism from consumer advocacy groups.

Editorial Team7/28/2026Updated 7/28/2026

AI Electricity Demand Pushes Grid Costs to Record Highs in PJM Auction

PJM Interconnection, operator of the largest regional power grid in the U.S., recently announced the results of its capacity market auction, with total costs reaching $16.4 billion. Of this, $6.3 billion—approximately 38%—was attributed to new electricity demand from AI data centers. This figure underscores the immense pressure that rapid AI industry growth is placing on power infrastructure and has become a key factor in the U.S. government’s energy policy adjustments.

The PJM grid covers 13 states in the eastern U.S. and Washington, D.C., serving over 65 million customers. This summer, the region faced a dual challenge of extreme heat and surging electricity demand from AI data centers, pushing grid loads to near-historic highs. In response, the U.S. Department of Energy issued an emergency order requiring large data centers to switch to on-site power systems when necessary to prioritize residential and medical electricity needs.

Companies Seek Independent Power Solutions to Reduce Grid Dependence

Facing grid supply pressures, some technology companies are exploring more autonomous energy solutions. Oracle plans to deploy approximately 2.5 GW of fuel cells at its Project Jupiter AI data center to enhance power stability. Meanwhile, AI infrastructure provider Crusoe has adopted natural gas power generation combined with microgrid technology to shorten the wait for grid connections and improve energy independence.

These moves reflect industry concerns over insufficient grid capacity. While most AI data centers remain connected to the public grid as a backup or supplementary power source, advancements in energy technology may lead more companies to shift toward self-built energy systems, with the public grid serving as a secondary source. This could even pave the way for more autonomous AI energy systems in the future.

Trump Administration Policy Shifts Costs to Data Centers, Drawing Consumer Group Criticism

To address rising electricity prices driven by AI data center demand, President Trump recently expanded the Ratepayer Protection Pledge, adding over 200 utility companies, data center developers, cooperatives, and state governments. Under this policy, large data centers will bear the costs of new power generation facilities and infrastructure rather than passing them on to general ratepayers.

However, the policy has faced skepticism from consumer advocacy groups. They argue that the Ratepayer Protection Pledge lacks legal binding power and enforcement mechanisms, making it difficult to ensure data center operators fully comply with cost-bearing principles. Critics also contend that the Trump administration’s encouragement of fossil fuels like natural gas to power AI data centers could increase long-term energy costs and environmental burdens, exacerbating climate change risks.

Electricity Price Pressures Emerge as EIA Forecasts Significant Increases Over Next Two Years

The U.S. Energy Information Administration (EIA) projects that, excluding inflation, average residential electricity prices in the U.S. will rise by 5.1% in 2026 and an additional 2.4% in 2027. This forecast reflects growing electricity demand and infrastructure investment pressures, though the EIA has not explicitly linked AI data center demand to the price increases.

The strain on the PJM grid highlights the limitations of traditional power systems in meeting the demands of emerging technologies. As AI applications continue to expand, electricity costs and environmental impacts will become critical considerations in industry development. Governments and businesses are actively seeking more flexible energy solutions to balance industrial growth with public needs.

Previous

Renewables Account for Over 30% of U.S. Electricity Generation for the First Time as Solar Power Output Surpasses Coal and Wind in May

Next

New York Imposes First-in-Nation Moratorium on AI Data Centers: Experts Say Environmental Debate Masks a Broader Battle Over AI’s Role in Society

0
0

Log in to comment and like articles.

LoginRegister

Comments

No public comments yet.

Back to articles