An Overly AI-Dependent Economy?
Analyzing whether modern US economic growth relies too heavily on technology sector investments.
1Headline GDP Masks Fragility
While current US real GDP continues to expand at a 3.4% annualized rate, the underlying composition of this growth is shifting. Recent data indicates that a disproportionately large share of economic activity now stems from information-processing equipment and software. This concentration raises concerns about the fundamental robustness of the broader American economy.
2Contrasting Two Economic Eras
Comparing the 1995–1999 expansion to the eight quarters ending September 2026 reveals a significant divergence. Late-1990s growth averaged 4.17% annually with broad participation, whereas the current period averages 2.34%. Today’s economy is growing more slowly and relies on a much narrower range of economic activity than the late-twentieth-century boom period.
3Technology Investment Dominance
In the late 1990s, information-processing equipment and software contributed 0.68 percentage points to quarterly annualized GDP growth. Today, that contribution has risen to 0.81 percentage points. Measured relative to total growth, technology-related investment now accounts for 35% of expansion, a massive increase from the 16% share observed during the late-1990s period.
4The Ex-IT Economy Decline
Subtracting technology-related contributions from GDP growth highlights a stark decline in other sectors. The non-IT portion of the economy contributed 3.49 percentage points during the 1990s, compared to just 1.53 percentage points recently. This represents a 56% reduction in the growth contribution from the rest of the economy, signaling a narrowing foundation.
5Wobbly Economic Pillars
Beyond technology, other traditional growth drivers are struggling. Residential investment, which contributed positively in the late 1990s, has turned negative during the latest eight quarters. High home prices relative to income and locked-in low mortgage rates are suppressing housing inventory and transaction volumes, further weakening the structural support for current economic expansion.
6AI Capital Expenditure Risks
Current spending on compute capacity, data centers, and software is driven by expectations of future productivity gains. If these expected financial returns disappoint, businesses may scale back investments rapidly. Such a reduction would directly weaken GDP and could trigger negative ripple effects across employment, construction, and consumer spending sectors beyond technology.
7Historical Context of Speculation
The late 1990s were not immune to speculative excess, eventually leading to the dot-com bubble burst and the 2001 recession. While today’s corporations and banks may possess different financial cushions, the historical comparison remains instructive. Even during that tech boom, the rest of the economy provided stronger, more diverse contributions than seen today.
8Implications for Future Stability
The current expansion is not necessarily signaling an imminent recession, but it is undeniably more concentrated. A healthy economy requires multiple powerful engines to sustain organic growth. Because the US has become overly dependent on a single driver, any reversal in technology investment could prove especially consequential for the broader national economic outlook.
Original article: The Big Picture — An Overly AI-Dependent Economy?