Alphabet Raises the Stakes on AI Infrastructure Spending

On Wednesday, Alphabet said it expects 2026 capital expenditures of $175 billion to $185 billion, a figure that would be more than double what the company spent in 2025.

While Alphabet beat Wall Street expectations on fourth-quarter revenue, earnings, and cloud growth, the size of the capex forecast weighed on shares in extended trading.

The spending will primarily go toward expanding AI compute capacity for Google DeepMind, scaling Google Cloud infrastructure to meet surging customer demand, and funding “strategic investments in other bets,” according to CFO Anat Ashkenazi.

Alphabet had previously warned investors to expect a “significant increase” in 2026 capex, but the magnitude of the forecast now exceeds projections from its hyperscaler peers, effectively resetting the bar for the industry.

Outpacing the hyperscalers

While competitors are also spending aggressively, Alphabet’s numbers stand out. Meta has guided to $115 billion to $135 billion in 2026 capex, nearly doubling last year’s total.

Microsoft declined to provide a full-year figure but said spending will decline sequentially after $37.5 billion in the most recent quarter. Analysts expect Amazon to lift capex to roughly $146.6 billion this year, up about 18%, according to FactSet.

Alphabet’s outlook lands well above those estimates — at a moment when investors are increasingly sensitive to AI-related spending. Despite solid earnings across Big Tech, the software sector has lost roughly 30% of its value over the past three months, reflecting fears that AI could disrupt existing business models and make massive infrastructure investments riskier.

Demand is driving the spend

Executives made clear the spending is demand-driven. Google Cloud’s backlog jumped 55% sequentially and more than doubled year over year to $240 billion at the end of the fourth quarter, Ashkenazi said. Cloud revenue rose nearly 48% from a year earlier, underscoring why Alphabet is racing to add capacity.

“The vast majority of our capex was invested in technical infrastructure,” Ashkenazi said, noting that in the fourth quarter roughly 60% went to servers and 40% to data centers and networking equipment — a mix that is likely to persist.

CEO Sundar Pichai emphasized that compute availability, not demand, is the binding constraint. Asked what keeps him up at night, Pichai’s answer was simple: “Compute capacity.”

“Be it power, land, supply chain constraints — how do you ramp up to meet this extraordinary demand for this moment?” he said.

AI momentum — and mounting costs

Alphabet highlighted several AI wins during the quarter. Its flagship Gemini app now has 750 million monthly active users, up from 650 million last quarter. Pichai also reiterated the importance of Google’s partnership with Apple, which selected Google as a preferred cloud provider as it overhauls Siri using Gemini models.

The company is also moving beyond organic buildouts. In December, Alphabet agreed to acquire data center operator Intersect for $4.75 billion, adding ready-made capacity amid tightening supply.

Internally, the pressure is even clearer. Amin Vahdat, who oversees AI infrastructure, told employees that Google needs to double serving capacity every six months to keep up with demand.

“The competition in AI infrastructure is the most critical — and the most expensive — part of the AI race,” Vahdat said.

For investors, Alphabet’s message is stark: AI leadership now requires capital at a scale that would have been unthinkable just a few years ago.

The open question is not whether Alphabet can spend this much — but whether the returns arrive quickly enough to keep Wall Street on board.

Disclaimer