The artificial intelligence boom is showing up in an unexpected place, the household electric bill, and financial researcher Jim Rickards says the story runs deeper than power rates. In a new presentation, he follows the strain on America's grid down to the metals required to build it. Peer reviewed research from North Carolina State University and partner institutions projects that electricity demand from data centers could raise U.S. power costs by a national average of 6%–29% by 2030 and by up to 57% in some regions. While some utility-backed analyses contest whether data centers are the main driver, the potential for significant cost increases has placed household bills squarely in the debate.

Lawrence Berkeley National Laboratory estimates data-center demand could reach 6.7%–12.0% of total U.S. electricity consumption by 2028, up from about 4.4% in 2023. Meeting that demand requires substantial new grid infrastructure. The metal that grid runs on is already tightening. Morgan Stanley forecasts a roughly 600,000 tonne refined copper deficit in 2026, the largest in more than two decades. This deficit stems from mine disruptions and limited new supply colliding with demand that S&P Global projects could rise 50% by 2030. Rickards uses this supply-demand gap to argue that the constraint on the AI build-out may be physical, not financial.

Supply Constraints and Market Parallels

Rickards draws a direct comparison to a situation that played out during Trump's first term. Weeks before the 2024 election, a plan to privatize Fannie Mae began circulating. The company had been bailed out by the government after the 2008 financial crisis and shares had collapsed, trading for a fraction of what the business was actually worth. Rickards says he recommended shares to a group of his readers before the news broke wide. In the year that followed, shares climbed more than 1,000%.

His point is not that history repeats exactly, but that the setup rhymes. When a real asset is being held back by a government decision rather than any flaw in the underlying business, the market tends to misprice it heavily until that decision changes. Rickards believes the copper sitting inside this one blocked American deposit is in the same position today. The asset is real, the demand is growing, and the only thing holding the price down is a regulatory decision he expects to shift.

Projected Impact on Consumption and Supply

The following table outlines the projected increases in electricity consumption and the forecasted deficit in copper supply.

Implications for Investors

For an investor, the situation reframes the obvious question, not "which AI company wins," but "what does the whole build out physically require, and who supplies it?" You cannot move that much new electricity without enormous quantities of copper, and a build-out of this scale runs straight into the question of where that copper comes from. The AI story everyone is watching may depend on a metals story almost no one is watching, and the companies that supply those metals, along with the cost of power itself, touch ordinary households and portfolios alike.