AI Spending Hits 45% of S&P 500 Capex, Shielding US Economy
Tech giants' record-breaking investment in data centers and AI hardware now surpasses dot-com era peak, driving economic growth and propping up corporate earnings.
A torrent of spending on artificial intelligence infrastructure has pushed technology's share of U.S. corporate investment to its highest level in history, a trend credited by analysts with single-handedly shielding the American economy from a potential recession.
Investment in AI and related technology now accounts for a record 45% of all capital expenditures across the S&P 500, according to data from Topdown Charts and LSEG. This unprecedented capital allocation, primarily for building out massive data centers and purchasing specialized processors, surpasses the roughly 39% peak seen during the dot-com bubble in 2000 and signals a fundamental rewiring of the U.S. economy.
The scale of this investment is staggering. In the third quarter of 2025 alone, Big Tech's capital expenditure on AI and data centers surged to $113.4 billion, a 75% increase from the previous year. This spending is highly concentrated among a handful of "hyperscaler" companies. Tech giants like Microsoft, Google, Amazon, and Meta are responsible for the vast majority of this outlay as they race to build the computational power necessary to train and deploy advanced AI models.
Microsoft, with a market capitalization of approximately $3.6 trillion, has been a key driver of this trend, investing heavily in its Azure cloud platform to support its partnership with OpenAI. This aggressive spending has had a direct impact on corporate earnings, underpinning much of the S&P 500's blended earnings growth of 13.1% year-over-year for the third quarter.
This capital tsunami is creating a distinct class of beneficiaries. NVIDIA Corporation, which designs the graphics processing units (GPUs) essential for AI, has seen its valuation soar to over $4 trillion. The company's recent quarterly revenue growth of over 62% reflects its dominant position as the primary supplier for the AI arms race. The impact also ripples through the supply chain to companies like Super Micro Computer, which provides the high-performance servers and hardware needed to house these powerful chips.
The macroeconomic implications are profound. Research from S&P Global recently concluded that the sheer scale of data center investment is now significant enough to "move the U.S. macro needle," directly contributing to GDP growth. This spending on physical infrastructure—construction, networking, and energy—provides a powerful economic stimulus that has, so far, offset weakness in other sectors.
Looking ahead, the trend shows no signs of slowing. Global data center spending is projected to exceed $500 billion in 2025, indicating a sustained, long-term investment cycle. However, the scale and speed of this capital shift are raising concerns. A recent Bank of America survey found that 45% of global fund managers now view an "AI bubble" as the most significant tail risk to the market. For now, the AI build-out is not just a technological revolution but the primary engine of the U.S. industrial and economic landscape.