$700B of AI Spending Has Moved From Equity Hype Into Credit Risk

$700 billion of projected AI infrastructure spending is large enough to affect borrowers, bondholders, and the wider economy.

Fitch estimated that 2026 capital expenditure by four major technology companies could rise more than 75% to about $700 billion. It also highlighted $182 billion of investment-grade bond issuance from a wider group of AI-linked companies and warned that an AI market correction is becoming a major global credit risk.

This is a different argument from “AI stocks look expensive.” Credit risk appears when companies finance huge projects on the assumption that future cash flows will cover interest, depreciation, power, and replacement costs. If AI revenue arrives slowly, the spending still leaves real obligations behind.

The boom currently supports growth. Fitch estimated IT investment added about 1.4 percentage points to first-quarter U.S. GDP growth. That creates a two-way macro risk: continued construction supports suppliers and employment, while a sudden pullback could hit data centers, utilities, chips, private credit, and regional economies at once.

The most revealing signal will be the gap between AI-related revenue growth and the total cost of serving it. Bond spreads, free cash flow, utilization rates, and customer concentration may say more than another model benchmark.

Not Financial Advice. Technology stocks and corporate bonds can lose value, and forecasts may change.

Is the AI buildout creating durable infrastructure, or pulling future demand forward with too much debt attached?