Google will double its CAPEX in 2026 as AI race intensifies

Alphabet (GOOGL) says capital spending in 2026 will land between $175 billion and $185 billion.

That’s nearly double last year’s $91 billion and far above the $115 billion Wall Street expected.

This is a declaration that the AI race has moved into an infrastructure war.

Most of the money goes into servers, data centers, and networking. That’s the hard metal needed to train and serve large AI models at scale. CEO Sundar Pichai admitted what everyone in the industry knows but rarely says out loud. Alphabet is compute constrained and will stay that way all year.

The spending lines up with accelerating demand. Cloud revenue jumped 48 percent to $17.7 billion in the fourth quarter, the fastest growth in more than four years. Gemini 3's sales are exploding. Eight million paid licenses across 2,800 companies isn’t a demo phase.

The stock is up 76 percent since early 2025, even after a small after hours dip. Investors are nervous about returns. Alphabet is telling them returns come later. Capacity comes first.

The disruption behind the news: This is about locking in who controls AI supply for the next decade.

Training frontier models now costs hundreds of millions per run.

Inference costs scale with every user query.

That means whoever owns the cheapest, densest compute wins.

Alphabet is choosing to buy that advantage outright rather than rent it over time.

Big Tech is on track to spend more than $500 billion on AI infrastructure this year. That level of capex creates a cost curve smaller players can’t touch. Startups won’t compete on raw model scale. Enterprises won’t build their own data centers. They’ll buy access from whoever already paid the bill.

Gemini’s consumer app now touches 750 million monthly users. AI Mode usage inside Search has doubled. Those interactions aren’t just product features. Every use makes the AI better and generates revenue for Google at the same time.

The Apple deal matters too. Apple doesn’t partner unless it has to. When Apple rents Alphabet’s cloud to support AI on iPhones, it’s an admission that even the richest hardware company alive doesn’t want to shoulder this compute load alone.

This spending wave also reshapes competition. Meta just raised AI capex 73 percent. Microsoft is posting record quarterly spend. The arms race is narrowing to three or four platforms that can fund expensivbe AI infrastructure for years.

What to watch next

Watch margins before profits.

Alphabet will look expensive and inefficient on paper as depreciation hits earnings. That’s normal. The tell will be cloud operating margins stabilizing while revenue keeps growing above 40 percent.

Watch enterprise lock in.

Eight million paid Gemini seats is step one. The next milestone is standardized AI workflows that make switching providers painful and slow.

Watch regulation lag reality.

Governments will talk about AI rules while compute ownership quietly centralizes. By the time regulators react, the physical infrastructure will already be in place.

Google CAPEX doubling is a sign that Alphabet is buying time, scale, and leverage. Anyone who thinks this money comes back fast is missing the point. This is how monopolies are built om real time.

Google (GOOGL) has a Disruption Score of 4. Click here to learn how we calculate the Disruption Score. 

Google is also part of the Disruption Aristocrats, our quarterly list of the world’s top disruptive stocks.