Microsoft, Alphabet, Amazon, Meta, and Oracle have collectively guided to $660–690B in capital expenditure for 2026, up from approximately $380B in 2025, with the majority directed at AI compute, data centers, and networking. All five report supply-constrained rather than demand-constrained markets; Microsoft disclosed an $80B Azure backlog unfulfillable due to power availability. Pure-play AI vendor revenues (OpenAI at ~$20B ARR, Anthropic at ~$9B run rate, all others combined likely under $35B) remain a small fraction of the infrastructure being built on their behalf, creating a meaningful gap between investment timing and revenue realization. For financial institutions, the key risk to monitor is whether the 18–36 month payback cycle on this infrastructure cohort holds if inference efficiency gains or adoption pace diverge from current hyperscaler assumptions. Also reported by: https://know.creditsights.com/insights/technology-hyperscaler-capex-2026-estimates/
559 other tracked developments touching the same claims.