$650 Billion on AI Capex

Quick Overview

Big Tech companies, following strong earnings reports from Google and Amazon, are projected to spend a massive $650 billion on AI infrastructure in 2026, driven by intensifying competition, which is causing concern in financial markets regarding the shift from stock buybacks to massive capital expenditures (Capex).

Key Points: Big Tech (Meta, Google, Amazon, Microsoft) is forecasted to spend approximately $650 billion on capital expenditures (Capex) in 2026, primarily for AI infrastructure like new data centers. Google reported annual revenue surpassing $400 billion for the first time, with its AI division driving demand, and its Gemini app reaching 750 million monthly active users by February 2026. Amazon's stock dropped after it announced plans to spend $200 billion on AI data centers and equipment in 2026, with CEO Andy Jassy defending the spending as an investment that will yield strong long-term returns, especially for AWS. Meta's full-year Capex is projected to rise to $135 billion (an 87% jump), while Microsoft is expected to spend almost $105 billion in Capex for its fiscal year ending in June. A key concern in financial circles is the shift from prioritizing stock buybacks to massive AI Capex, which drains liquidity from the financial system, as noted by analyst commentary referencing Stanley Druckenmiller's concerns. OpenAI announced its new platform, Frontier, designed to help enterprises build, deploy, and manage AI agents securely, with early adopters including HP, Oracle, State Farm, and Uber.

Context: This news briefing summarizes recent major earnings reports and strategic spending announcements from leading technology companies—Google, Amazon, Meta, and Microsoft—focused heavily on massive capital expenditure (Capex) dedicated to building out Artificial Intelligence (AI) infrastructure. The discussion contrasts this investment trend against the previous decade's focus on stock buybacks, highlighting market concerns about liquidity and the scale of required AI spending.