Goolsbee Warns AI Data-Center Boom Is Bringing the Economy Close to Overheating

AI infrastructure spending is pushing the U.S. economy toward an overheating risk, Chicago Federal Reserve President Austan Goolsbee said in an interview with Fortune. He described data-center expansion as “very hot” and said demand for construction, power, equipment and other resources is competing with activity elsewhere in the economy.

The concern is economic allocation, not simply enthusiasm for new software. When data centers absorb scarce workers, electricity, financing and materials, other businesses face higher costs or reduced access to resources. Goolsbee characterized the result as AI growth “shoving other parts of the economy down,” a warning that a celebrated investment boom can create broader inflation pressure even while raising productivity in selected industries.

Broad consumer spending, rather than AI data-center construction, has kept the economy stable, Goolsbee said. Employment indicators including unemployment, vacancies, hiring and layoffs remain relatively steady, giving the Federal Reserve room to watch how demand and inflation develop.

That stability also raises the stakes around AI investment. A strong consumer sector combined with unusually heavy infrastructure spending creates two sources of demand at once. Goolsbee told Fortune that the real economy had moved from stability toward overheating risks, while progress on inflation stalled and then worsened after several years of improvement. CNBC separately highlighted persistent energy inflation as a continuing pressure on prices, linking the AI buildout to the wider cost environment through power demand and infrastructure needs.

The key question for monetary policy is whether AI spending produces enough productivity growth to justify its inflationary costs. If data-center construction and related purchases lift output quickly, the investment can expand supply and offset some price pressure. If spending rises without comparable productivity gains, demand grows faster than the economy’s capacity.

That distinction underpins the rate risk described in Goolsbee’s comments. Finance.yahoo reported that the Fed could need to raise rates if the AI boom drives spending without delivering sufficient productivity. Higher rates would cool investment, housing and consumer demand, while also raising the cost of financing the data centers that power AI services. The resulting policy challenge is unusually concentrated: a boom in one capital-intensive sector can keep headline growth strong while crowding out smaller firms and other forms of investment.

Energy costs are an immediate concern because data centers require large and continuous electricity supplies. Persistent energy inflation compounds the effects of construction demand, equipment orders and competition for skilled labor, making the AI buildout relevant to the Fed’s price-stability mandate even before its effects on productivity become clear.

Goolsbee’s broader assessment places AI within an economy that has already lost some of its inflation momentum. The labor market remains broadly stable, but inflation readings have delivered mixed signals. That combination leaves policymakers balancing two risks: tightening too early and weakening ordinary consumer activity, or waiting too long while infrastructure demand and energy costs spread through the economy. Bloomberg’s account of Goolsbee’s overheating concerns and CNBC’s focus on energy inflation point to the same policy pressure from different angles.

AI companies face greater scrutiny over whether capital spending translates into measurable output, revenue and efficiency gains. Data-center operators, chip suppliers, utilities and cloud providers are tied to the investment cycle, so a shift toward tighter monetary policy would affect the entire buildout rather than a single model developer.

The immediate implication is a stronger demand for evidence. Firms expanding capacity will need to show that new computing power supports usable products, higher productivity or durable customer demand. Investors and policymakers will also watch whether infrastructure spending pulls resources from manufacturing, energy-intensive businesses and consumer services. The Stanford Institute for Economic Policy Research framed the issue as a question of whether AI exuberance can overheat the economy, while Goolsbee’s remarks supply the central policy concern: growth in AI infrastructure must be measured against its demands on the rest of the economy.

The Federal Reserve’s response will depend on incoming inflation, spending, productivity and labor-market data. Goolsbee’s comments do not announce a rate decision, but they establish that AI investment has entered the Fed’s macroeconomic risk assessment alongside consumer demand and energy prices.

A sustained rise in spending without matching productivity would strengthen the case for restrictive policy. Conversely, clear efficiency gains could help absorb the infrastructure boom by expanding the economy’s productive capacity. For now, the warning is measured but direct: AI data centers are growing fast enough to compete with other sectors for resources, and the Fed is watching whether that competition becomes a broader inflation problem.