The AI Buildout Is Fueling The Stock Market. It Could Also Be Contributing To Keeping Inflation Stubborn.

An argument in favor of the artificial intelligence boom is that it could make businesses dramatically more productive and goods and services cheaper. For the Federal Reserve, however, AI is creating a more immediate problem in that area.

The enormous cost of building the infrastructure needed to power it may be adding to inflation before those promised productivity gains arrive. Silicon Valley leaders including OpenAI CEO Sam Altman and Elon Musk have predicted that increasingly capable AI and robotics will ultimately push costs sharply lower.

Altman has argued that “intelligence too cheap to meter” is within reach, while Musk has forecast an era of abundance driven by automation. But according to CNBC, the economic effects visible today look considerably different.

Companies are pouring hundreds of billions of dollars into data centers, chips, electricity and other infrastructure while adoption across the broader economy remains uneven. That means many of AI’s costs are arriving well before its potential benefits.

Goldman Sachs Research estimates U.S. capital spending on the AI build-out will reach $581 billion this year, equivalent to about 1.8% of gross domestic product. Globally, investment could reach $1 trillion. By 2028, Goldman estimates U.S. AI investment could represent 2.8% of GDP.

At the same time, only about 17% to 20% of U.S. businesses reported using AI, according to a Census Bureau survey published in May. Adoption is also much higher among large companies than smaller businesses. “For it to impact the economy, it has to be adopted by organizations,” OpenAI chief economist Ronnie Chatterji told CNBC. “Those organizations have to realize value.”

Chatterji said it could still take time before AI’s effects become clearly visible in productivity statistics. OpenAI’s own data also points to a widening gap between companies aggressively reorganizing their operations around AI and those using it more casually.

That lag matters enormously for the Fed. The central bank is trying to determine whether AI should be treated primarily as a future disinflationary force or a current source of additional demand and higher prices. Fed Chairman Kevin Warsh has argued that AI could ultimately increase productivity, strengthen U.S. competitiveness and put downward pressure on inflation.

But other policymakers are more concerned about what is happening now. Fed officials voted in July to keep the benchmark interest rate between 3.5% and 3.75%. Minneapolis Fed President Neel Kashkari dissented in favor of a higher rate, warning that massive investment in data centers was creating “a new demand element” contributing to inflation.

One of the clearest pressure points is electricity. AI data centers require enormous amounts of power, adding demand to an electrical grid already facing infrastructure constraints. Household electricity prices increased 10% during the two years through July, according to Bureau of Labor Statistics data cited by CNBC. Overall consumer prices increased 6.2% during the same period.AI’s infrastructure boom is also straining technology supply chains.

Demand for advanced servers and Nvidia chips has surged as technology companies race to expand computing capacity. Memory prices are another concern. JPMorgan Chase estimates the price of dynamic random access memory, or DRAM, will have increased 400% by the end of 2026 compared with 2024.

Meanwhile, the consumer price index shows prices for computer software and accessories have risen 22.4% since July 2024. There is also uncertainty over just how transformative AI will ultimately prove for productivity.