AI Spending Just Outpaced Big Tech's Earnings - And It's Never Happened Before

There's one chart every investor should be looking at right now. It shows two lines - Big Tech's capital expenditures and their free cash flow. For years they tracked each other closely, with FCF always running ahead. Then, in Q1 2026, something happened that had never happened before: the lines crossed.

AI Capex - $133B. Free Cash Flow - $119B. For the first time ever, Big Tech is spending more on artificial intelligence than it actually earns.

How We Got Here​

A year ago, everything still looked manageable. In Q1 2025, the combined FCF of the six largest hyperscalers - Microsoft, Amazon, Google, Meta, Apple, and Nvidia - sat at around $95B, while capital expenditures came in under $76B. There was a comfortable buffer, shareholders were getting buybacks, and nobody was asking hard questions.

Then the acceleration began. Capex climbed quarter after quarter:

• Q1 2025: $76B capex vs $95B FCF — comfortable gap
• Q2 2025: $93B capex vs $91B FCF — buffer nearly gone
• Q3 2025: $102B capex vs $102B FCF — lines converge
• Q4 2025: $123B capex vs $134B FCF — brief relief
• Q1 2026: $133B capex vs $119B FCF — lines cross

That Q4 spike in FCF created a false sense of security. Then Q1 2026 arrived, FCF pulled back to $119B, capex kept climbing - and the crossover became reality.

What This Actually Means​

When a company spends more than it earns in free cash flow, something has to give. The options are straightforward: take on debt, draw down balance sheet cash, or cut other spending - like share buybacks. Based on the data, Big Tech has been doing all three at once.

This isn't a crisis. These companies carry trillion-dollar balance sheets and can access cheap debt whenever they need it. But it is a fundamentally new situation. Until now, AI infrastructure was being built from surplus. Now it's being built at the limit - and in some cases, beyond it.

The Question Nobody Can Ignore​

All of this would make perfect sense if AI were already generating revenue on a scale that justified the investment. But it isn't — not yet. These companies are betting that monetization will catch up to capex. Maybe it will. But right now, the market isn't funding profit. It's funding the promise of profit.

And the longer those two lines keep moving in opposite directions, the louder one question becomes: when will AI start generating returns as aggressively as it's being invested in?