Big Four Hyperscalers Raise AI Capex Targets to $745 Billion Amid Surging Demand
Amazon, Alphabet, Meta, and Microsoft boost 2026 AI infrastructure spending forecasts, signaling an accelerated AI arms race driven by enterprise demand.
The story
The four major U.S. hyperscalers—Amazon, Alphabet, Meta, and Microsoft—have increased their combined capital expenditure forecast for this year to a range of $720 billion to $745 billion, up from a previous projection of $695 billion to $725 billion. This aggressive investment expansion is a response to surging enterprise AI demand, with Google Cloud notably reporting an 82% jump in revenue.
Amazon raised its annual capex forecast the most, from $200 billion to $220 billion, while Google lifted its range to $195 billion–$205 billion, and Meta adjusted its lower guidance to $130 billion. The market is closely observing how quickly these substantial investments will translate into tangible profits for the companies.
Who moved
Nvidia
What Changed: Partnered with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish a $500 billion independent computing power financing platform.
Why It Matters: This initiative aims to provide funding support to Nvidia customers for AI infrastructure construction at attractive rates, potentially accelerating AI adoption.
Meta
What Changed: CEO Mark Zuckerberg outlined a "personal super-intelligence" roadmap, proposing to deliver super-intelligence to individuals and small businesses for free or low cost, and open-sourcing some models.
Why It Matters: This details Meta's long-term vision for AI product development and its approach to AI governance and accessibility.
What Changed: Consolidated its AI leadership by appointing Koray Kavukcuoglu to run its combined AI research and operations, with Demis Hassabis becoming chairman of Google DeepMind and Alphabet's Chief Scientist.
Why It Matters: This structural move indicates Google's preparation to compete more aggressively in the AI space.
Microsoft
What Changed: Is in talks with TSMC to reserve capacity for producing over 300,000 of its upcoming Maia 300 AI processors for delivery in 2027.
Why It Matters: This move signals Microsoft's efforts to enhance its custom chip capabilities and reduce reliance on third-party silicon.
Products & launches
Muse Glimmer
Company: Meta
What: Open-sourced a 30B-parameter agent model under the Apache 2.0 license, capable of running locally on devices with 24 GB of VRAM and optimized for agentic scenarios.
Platforms & policy
Mozilla
Development: Filed a new friend-of-the-court brief on August 9, 2026, warning that banning payments from Google to search engine distributors would dramatically reduce its revenue if forced to replace Google with Bing as Firefox's default.
Money & markets
Alphabet. Raised its 2026 capital expenditure guidance to a range of $195 billion to $205 billion, an increase from its prior forecast of $180 billion to $190 billion, reflecting continued heavy investment in AI infrastructure.
Amazon. Increased its annual capital expenditure forecast from $200 billion to $220 billion, indicating a significant acceleration in its AI and cloud infrastructure buildout.
Meta. Adjusted the lower end of its annual capital expenditure guidance from $125 billion to $130 billion, reflecting ongoing investments in its AI initiatives.
Google Cloud. Reported an 82% year-over-year revenue growth in the second quarter, reaching $24.8 billion, with operating income more than tripling to $8.8 billion.
What we'll be watching
- Nvidia's second-quarter earnings report on August 26.
- The upcoming July Consumer Price Index (CPI) report, which could influence Federal Reserve interest rate forecasts.
- Google is expected to file new arguments with the appellate court next month regarding antitrust penalties.
Reporting + analyst voices: grounded via Google Search at publish time.