Big Tech's AI spending is hitting levels you rarely see in modern history. $Amazon.com(AMZN)$ $Alphabet(GOOGL)$ $Meta Platforms, Inc.(META)$ $Microsoft(MSFT)$ $Oracle(ORCL)$ are estimated to push CapEx to 2.4% of U.S. GDP in 2026, up roughly 1 percentage point from 2025.
For some context, broadcasting and telecom companies spent about 1.0% of GDP on CapEx back in 1999, right at the peak of the Dot-Com investment boom. And the spending isn't slowing. By 2027, Big Tech CapEx is projected to reach 3.1% of GDP, more than double the 1.4% recorded in 2025.
In 2024, Big Tech CapEx was just 0.8% of GDP—roughly equivalent to telecom spending in 1997. The takeaway is pretty clear: the AI investment cycle is already surpassing the capital-spending boom of the Dot-Com era.
The bigger question now is whether AI-driven productivity and revenue growth can actually justify the scale of this investment.
Comments