Google’s $175–185B AI Bet

Alphabet is preparing to spend $175 billion to $185 billion in 2026, one of the biggest capital programs yet in the artificial intelligence race. (abc.xyz) Alphabet gave that forecast on February 4, 2026, during its fourth-quarter earnings call, after reporting $91.45 billion in capital spending for 2025. Chief Financial Officer Anat Ashkenazi said the money would go to servers, data centers and networking equipment for Google DeepMind, Google Cloud and other bets. (abc.xyz) At Google Cloud Next in Las Vegas on April 22, Chief Executive Sundar Pichai repeated the spending plan and said “just over half” of Alphabet’s machine-learning compute investment in 2026 is expected to go to the cloud business. Reuters reported the conference ran for three days and centered on selling AI tools to large companies. (blog.google) (usnews.com) Google’s pitch is that AI agents are moving from demos to office software. Thomas Kurian, the head of Google Cloud, said companies are now asking how to manage “thousands” of agents, and Google used Next to launch a Gemini Enterprise Agent Platform with governance and security controls. (blog.google) (usnews.com) The scale of the buildout reflects how AI services work behind the scenes. Training and running models requires specialized chips, power-hungry server clusters and high-speed links between data centers, so the companies selling AI software are also pouring cash into the hardware underneath it. (cnbc.com) (usnews.com) Google says the demand is already showing up in usage. On April 22, the company said nearly 75% of Google Cloud customers use its artificial intelligence products, 330 customers processed more than 1 trillion tokens each over the past 12 months, and its first-party models now handle more than 16 billion tokens per minute through direct application programming interface use, up from 10 billion last quarter. (blog.google) (cloud.google.com) Alphabet is also trying to show that the spending is tied to a real business, not just research. In the February earnings call, Pichai said Google Cloud revenue grew 48% year over year to an annual run rate of more than $70 billion, backlog jumped 55% quarter over quarter to $240 billion, and Alphabet had sold more than 8 million paid seats of Gemini Enterprise four months after launch. (abc.xyz) Investors have not greeted every giant AI budget warmly. CNBC reported Alphabet’s spending outlook exceeded what peers had laid out, while Microsoft said its capital spending would decline sequentially in the current quarter and Meta projected $115 billion to $135 billion of 2026 spending. (cnbc.com) Google’s answer is to own more of the stack itself. At Cloud Next, Pichai highlighted eighth-generation Tensor Processing Units, Google’s in-house artificial intelligence chips, alongside new security products tied to Wiz and a broader push to package models, tools and controls under the Gemini Enterprise name. (blog.google) (usnews.com) That leaves Alphabet making a straightforward wager in 2026: if companies buy agents and AI services at scale, the winners may be the cloud groups that already own the chips, networks and data centers those systems run on. (cnbc.com) (usnews.com)