Big Tech's AI spending spooks investors

Summary

Capital expenditure plans exceeding $740 billion by major tech companies triggered a sharp sell-off as Alphabet, Amazon, and Meta all reported negative free cash flow. This indicates that spending on AI infrastructure is now consuming the cash flows meant to finance it. Alphabet’s 12% single-day drop, erasing $500 billion in market value, and the Nasdaq’s 2.9% decline served as visible proof that investor enthusiasm has turned into anxiety.

Why it matters

This situation is critical because AI infrastructure is no longer funded solely by operational profits. The top five companies borrowed $326 billion in 2026, triple the previous year’s figure. This means the bet on AI returns is becoming increasingly leveraged. If the technology driving spending (e.g., Google’s AI Mode) is also eroding the revenue base (search ad growth at only 12%), the cycle of 'spending to protect the business that funds the spending' shifts from a growth story to a genuine solvency question.

Highlights

• Alphabet reported $9.1 billion in negative free cash flow in the third quarter. Search ad growth fell to 12%, its lowest level in five quarters. The company’s own AI Mode feature is competing with the ad engine it relies on.
• Amazon’s free cash flow over the last 12 months dropped from +$14.8 billion to −$42.4 billion. It outlined a $220 billion capital expenditure roadmap. Despite AWS growing 39%, the company’s stock value fell 7%.
• Microsoft was the sole exception with +$18.6 billion in positive free cash flow. However, due to third-party capacity delays, capital spending fell below the company’s own $50 billion forecast. This unsettled smaller AI compute leasing companies.

Critical take

The article frames the sell-off as a rational repricing of risk. However, it ignores that similar 'crack' narratives existed during the 2000 telecom and 2014 cloud-hyperscaler infrastructure investments, both of which ultimately yielded returns. The question the article actually avoids is this: Is the $326 billion debt burden a temporary bridge, or a structural admission that AI’s unit economics have not yet closed?