The AI Tech Boom’s Secret: A K-Shaped Hole in the Global Economy

At the heart of the US economy, a divergence is brewing. As the world’s largest economy, boasting an unrivalled technology sector concentrated mainly around Silicon Valley, the United States has been the biggest beneficiary of the AI tech boom. On the surface, traditional metrics of economic growth seem strong. GDP is expected to grow a healthy 2.2% in 2026, largely driven by AI-powered investment providing strong productivity gains. Yet these headline figures increasingly obscure a more uncomfortable reality. The US is not experiencing one economy, but moving in different directions, the dividing line running directly through the source of the strong GDP performance – artificial intelligence.

Analysts have taken to describing the current situation in US financial markets of a “K-shaped” economy, where higher-income households and stronger industries continue to boom, whilst lower-income consumers, small businesses, and more vulnerable sectors lag behind. This divergence is driven by large technology and AI capital expenditure in the world’s largest economy, resulting in a significant overvaluation of AI stocks as purchasing power of the median consumer is increasingly squeezed. In 2025, behind the headline story of record highs hit by the S&P 500, about 40% of the index’s total gains came from a handful of tech stocks nicknamed the “Magnificent Seven”, concentrating wealth sharply at the top.

Meanwhile, the bottom branch of the “K” tells a completely different story. Job growth has virtually flatlined, believed to be a consequence of automation mostly concentrated among younger workers. Despite a stronger-than-expected January, 2025 job growth in the US was even weaker than expected. Business surveys further suggest that many companies are hesitant to spend big, with elevated interest rates, rising input costs, and policy uncertainty weighing heavily on their decision-making.

The implications of this trend extend well beyond America’s borders. The IMF notes that IT investment as a share of US economic output has provided a major boost to business investment, generating positive spillovers globally. But these spillovers carry a hidden risk. AI-related goods are powered by high-tech semiconductors, which made up nearly half of trade flows in early 2025. A sharp correction in US AI investment would expose the global economy to serious damage. The only net positive growth in US investment is in the tech sector, which the IMF fears is not sustainable. Even a moderate correction in AI stock valuations could reduce global growth significantly, with far-reaching consequences even in economies with little exposure to technology, due to higher external borrowing costs and reduced demand and consumption in other sectors.

The impact of a K-economy may be overstated. A study found that in San Francisco, for each new technology job created, around 4.4 additional jobs are created in adjacent sectors, such as retail, education, and healthcare. If that multiplier holds, productivity gains at the top could spillover and lift wages more broadly. In this case, AI will not so much as replace workers but instead create new opportunities that would otherwise not exist. However, this outcome is far from guaranteed – this particular example is confined only to a single city and offers limited usefulness to gauge a global response. IMF managing director Kristalina Georgieva warns of an “accordion of opportunities that is open to some and not others”, a clear signal the AI boom is widening the gap between the winners and the losers.

The K-shaped economy is, in one sense, nothing new. Technological revolutions throughout history have always created winners and losers and the gains from past waves of automation eventually diffused themselves across consumers. But history also offers a cautionary tale. The dotcom bubble promised a similar rewrite of the global economy as the AI boom, before a correction wiped trillions from the stock market. There are important differences this time, particularly as the “Magnificent Seven” stocks in the current US stock market boast far more impressive profitability and revenue models, but the stakes are much higher, and a globally-owned US equity market threatens a much harder landing. The speed of the AI tech boom likewise raises questions of whether this particular case is moving too fast for a diffusion to occur. If these productivity gains remain locked inside the most successful tech firms, the divergence of the two faces of the US economy will compound even further. And in a financially-interconnected world, this problem will not be confined solely to Washington.