The explosive growth of artificial intelligence and cloud computing is reshaping the US energy landscape. The resulting surge in electricity demand has become a major driver pushing up electricity prices, leading to widespread public complaints. Facing public pressure, the federal government is brewing a major move—collaborating with multiple state governments to introduce an unprecedented "Emergency Power Auction Mechanism," which directly requires tech giants to fund the construction of new power generation facilities. This is the first time that energy policy has explicitly intervened to alter the cost structure and responsibility sharing within the AI industry.

**Electricity demand soars, data centers become the biggest driving force**

PJM Interconnection, responsible for power dispatch in the Mid-Atlantic and Midwest regions of the US, serves over 67 million people. It is now facing unprecedented electricity pressure—especially in Northern Virginia, home to the world's largest data center cluster.

PJM's forecast data is alarming: by 2030, the system peak electricity demand will increase by 17% compared to this year's peak. However, the construction speed of new power capacity cannot keep up with demand growth. As a result, recent power auctions have been unusually tight, with supply shortages occurring frequently. In the latest auction, the missing power capacity alone reached 6.6 GW—and the frantic expansion of data centers is at the core of this energy crisis.

**Rising electricity bills for households, political pressure forcing policy changes**

The tight power supply has directly reflected in household bills. According to the National Energy Assistance Directors Association, the national average retail electricity price jumped 7.4% in September, setting a record. This increase has imposed real burdens on low- and middle-income families and has sparked widespread dissatisfaction.

Political realities have forced decision-makers to take action. The proposal for the Emergency Power Auction Mechanism essentially shifts the cost burden of AI infrastructure development from ordinary consumers to tech companies. This policy shift means that future energy costs will become a key variable in the AI industry’s competitive landscape—those who can access cheap electricity more efficiently will gain an advantage in this infrastructure race. For investors paying close attention to global technology and financial patterns, this trend is worth monitoring closely.