KEY POINTS
- Apollo Chief Economist Torsten Slok says AI-related spending will account for roughly half of U.S. economic growth in 2026.
- AI exposure now spans equity markets, corporate debt and infrastructure spending, creating new concentration risks for investors and the broader economy.
AI-related economic activity now accounts for approximately half of all U.S. economic growth in 2026, according to Apollo Global Management's Chief Economist Torsten Slok. That growth is being fueled by an unprecedented wave of capital spending by some of the world's largest companies. Google, Amazon, Microsoft, and Meta collectively plan to allocate $725 billion to capital expenditures in 2026, and JPMorgan estimates global AI investment could exceed worldwide military spending by 2027.
During an appearance on "The Real Eisman Playbook" podcast hosted by former Neuberger Berman money manager Steven Eisman, Slok estimated that spending on data centers, semiconductors, electric power infrastructure, and related construction combined will contribute about 1% of total U.S. GDP growth. Slok attributes the other half of the total 2% projected growth to industrial reshoring (0.3%) and the Trump administration's 'One Big Beautiful Bill' tax cuts (0.9%), which boosted average refunds from $3,000 to $4,000.
Eisman noted that half of all newly originated corporate debt in 2026 is AI-linked, and AI-related corporate debt now accounts for 15% of the entire corporate bond market.
The same AI dominance is evident in the stock market with technology and technology-adjacent companies like Amazon and Google constituting 50% of the S&P 500's total value according to Barron's.
This unprecedented concentration across stocks, corporate debt, and capital spending makes the broader economy dependent on the fortunes of not only a single industry but a single technology within that industry.
AI critics like journalist Ed Zitron have interpreted the circular character of AI financing as a sign that an asset bubble is brewing, with established tech companies investing in AI startups which then sign deals with the companies that invested in them.
Yet even if such bubble predictions turn out to be mistaken, the real risk is concentration.
If major AI companies fail to become profitable enterprises or expensive token purchasers, like Uber migrate from Anthropic or OpenAI to vastly cheaper Chinese models, that would create a potentially large ripple effect across the entire economy. Equity and bond markets would face disruptions and so would the energy and infrastructure sectors since they are tied to the AI buildout, not to mention niche asset classes like private credit. Even traditionally conservative portfolio strategies, such as owning stock indexes or a 60-40 stock-and-bond ratio, might find themselves challenged now that AI exposure spans asset classes, undermining the benefits of diversification.
