Alphabet announced Wednesday it expects to spend $175-185 billion on capital expenditure in 2026, essentially doubling the $91-93 billion spent in 2025. The projection exceeds forecasts from Microsoft and Meta, resetting the industry benchmark for AI infrastructure investment.
The spending will fund AI compute capacity for Google DeepMind, cloud infrastructure to meet customer demand, and strategic bets in other areas, according to CFO Anat Ashkenazi. Google Cloud drove the rationale: revenue surged 48% year-over-year to $17.7 billion, while backlog doubled to $240 billion.
Market reception: skeptical
Alphabet shares dipped in extended trading despite beating revenue and earnings expectations. The company reported $113.8 billion in quarterly revenue (up 18%) and $34.5 billion in profit, crossing $400 billion in annual revenue for the first time. Wall Street didn't care.
The software sector has lost 30% of its value in three months, according to CNBC data, reflecting investor concerns about AI spending efficiency and ROI timelines. Microsoft's stock dropped 6% on similar capex uncertainty. Meta's rose, suggesting investors differentiate based on perceived capital efficiency.
The regional angle
For APAC enterprise leaders, this spending trajectory has implications. Hyperscaler infrastructure capacity increasingly concentrates among well-capitalized incumbents. GPU availability, pricing power, and service differentiation will reflect these investment patterns.
The question isn't whether AI infrastructure requires massive investment - clearly it does. The question is whether customer demand at current price points justifies doubling down again. Alphabet's cloud backlog suggests yes. The stock market isn't convinced yet.
Notably, Alphabet attributed October's "significant increase" warning to customer demand. Three months later, that translated to a spending forecast exceeding all peers. Either demand accelerated dramatically, or initial guidance was conservative. We'll see which when Q1 results arrive.