AI Spending Now Drives More Than a Quarter of US GDP Growth

The US economy is now growing largely on the back of one thing: spending on artificial intelligence. A Bloomberg analysis published this week found that AI-related investment accounts for more than 25 percent of total US GDP growth, marking the largest contribution from a single spending category in the history of the American economy.

A Bigger Scale Than 2000

Bloomberg's analysts use the dot-com bubble of a quarter-century ago as their benchmark. Back then, spending on IT hardware, software, and research and development peaked at around 6.5 percent of US GDP before the market collapsed. Today, AI-related spending has already reached about 8 percent of GDP, exceeding that earlier record even though the current boom has only been underway for a few years.

The difference from 2000, however, lies in the financial health of the companies now investing the most. Unlike many internet firms of two decades ago, today's AI leaders, such as Microsoft, Alphabet, and Meta, generate billions in cash flow and are profitable. Optimists point to this as the reason the current scale of investment need not end in an identical crash.

Who Is Spending the Most

The biggest spenders are companies simultaneously building data centers, buying chips, and developing their own models. Bloomberg names five companies as the engine of this boom: Microsoft, Amazon, Alphabet, Meta, and Nvidia. Over five years, these firms have increased their combined debt obligations by $350 billion, largely to finance the buildout of AI infrastructure.

Makers of advanced chips, according to the analysis, are selling virtually every processor they can produce. Demand for AI chips has outstripped manufacturing capacity for months, keeping prices high and driving up revenue for memory and logic chip makers alike.

What Happens If the Boom Slows

Analysts cited by Bloomberg point to the risk on the other side of this trend. If companies decide they have already built enough data centers, or if the pace of AI adoption in business starts to slow, the US economy would lose what is currently its single biggest growth engine. The scenario echoes what happened after the dot-com bubble burst, when capital spending outran actual demand, and its slowdown dragged on economic growth even before the first internet companies collapsed.

An additional note of caution comes from data on the efficiency of these investments themselves. Cited estimates suggest that only about 18 percent of the tokens consumed by AI coding tools ultimately translate into deployed, working products. That means a significant share of costly computing resources isn't directly generating measurable business value, at least not yet.

What It Means for Poland

For Poland's economy, the scale of US AI investment matters indirectly but tangibly. Much of global demand for cloud services, chips, and computing capacity is driven precisely by spending from major American tech companies, which affects hardware prices, the availability of computing power, and the pace at which new AI models reach markets like Poland's. Any slowdown in this US investment cycle could ripple through the global cloud market and the price of AI services that Polish companies also rely on.

Polish economists and investors are also following this story for another reason: the share of AI spending in US GDP has become one of the central arguments in the debate over whether the current stock market rally, driven by tech companies, rests on solid fundamentals or resembles earlier speculative bubbles. The answer bears on, among other things, the valuation of pension funds and savings indirectly invested in US stock indices.

Sources: US Economy Is Hooked on AI Spending (parkiet.com), The U.S. Economy Is Addicted to AI Spending. What Happens If It Slows? (finance.yahoo.com)