Big Tech AI CapEx Wave Lifts Data Center Names
Shares in key data center and infrastructure-related names are trading higher today as fresh guidance from major cloud providers underscores a massive, accelerating build-out in AI infrastructure. The move reflects a broader market read that the AI capex cycle is still in early innings, with hyperscalers signaling multi‑hundred‑billion‑dollar commitments over the next two years.
What Happened
Recent earnings commentary and analyst notes highlight that the largest U.S. cloud and AI infrastructure providers—$MSFT, $GOOGL, $AMZN, $META, and $ORCL—are collectively on track to spend roughly $660 billion to $690 billion on capital expenditures in 2026, nearly double their 2025 levels. Bank of America and other research firms now project global hyperscale capex to reach around $611 billion this year, up sharply from 2025, as hyperscalers pour money into GPUs, servers, data centers, power, and cooling.
Within that, AI-specific infrastructure—dedicated compute, networking, and data center capacity for AI workloads—accounts for an estimated 75% of the spend, or roughly $545 billion of the projected $725 billion hyperscaler capex in 2026, according to CFA-aligned research released this week. Earlier this month, Alphabet alone raised its 2026 capex guidance to a range of $180 billion to $190 billion, with CFO Anat Ashkenazi flagging that 2027 capex is expected to "significantly increase" compared with 2026.
Analyst Take
Analysts are increasingly framing the AI capex surge as a structural tailwind rather than a short-term boom. Bank of America’s latest note describes AI data center investment as on track to triple to more than $1.2 trillion by 2030, while Goldman Sachs projects global AI capex to hit $539 billion this year, up 36% from 2025, and $629 billion in 2027. The bank notes that even with slower growth in 2027, the absolute dollar ramp remains substantial.
Several research desks are highlighting that the current hyperscaler build-out is outpacing the physical supply of data center capacity, which could extend pricing power and utilization for data center real estate investment trusts (REITs) and specialized providers. The combination of rising GPU density, power constraints, and cooling requirements is also seen as a long-term driver for infrastructure and energy-related plays that support AI data centers.
What to Watch
For investors, the next key data points will be quarterly earnings from hyperscalers and their capex disclosures, as well as updated analyst estimates on global AI and hyperscale capex. Any sign of a slowdown in the growth rate of AI-related capex, or a shift in hyperscaler guidance, could pressure data center and infrastructure names that have rallied on the capex narrative. Conversely, if hyperscalers continue to raise or reaffirm aggressive 2026 and 2027 capex targets, the AI infrastructure theme is likely to remain a central driver for cloud, data center, and select hardware and power infrastructure names.