
The tab just got fatter
Alphabet told the market it’s planning to spend even more on AI infrastructure in 2026, lifting its capex guidance to $195–205 billion. That’s not pocket change; that’s “did somebody order the entire server aisle?” money.
The company says the bump is about capacity shortages, which is corporate-speak for: customers want more, faster, and Alphabet’s trying to keep the trains running without turning the cloud into a waiting room.
Why investors aren’t freaking out
This isn’t just Alphabet tossing cash into the AI furnace for vibes. Google Cloud revenue jumped 82% to $24.8 billion, and its backlog hit $514 billion. In other words, the company isn’t only spending more — it’s also seeing enough demand to justify the shopping spree.
That matters because Alphabet has a weirdly powerful AI setup:
- Gemini gives it the model layer
- TPUs give it custom chips
- Cloud gives it the customer base
- Networking ties the whole machine together
So unlike some AI hopefuls that only make money if the model gods smile on them, Alphabet can potentially monetize the same investment across multiple businesses. Very efficient. Very Silicon Valley.
The big picture
The real question isn’t whether Alphabet is spending a ton — it’s whether that spending builds a moat or just a more expensive moat. Right now, the numbers say demand is strong, the infrastructure is busy, and the AI arms race is still very much on.
Big picture: when a company raises capex this hard and investors mostly nod along, it usually means the market thinks the runway is still long.
