Google Ramps AI Capex for Always-On Search Agents as Token Use Surges
Updated
Updated · The Register · May 19
Google Ramps AI Capex for Always-On Search Agents as Token Use Surges
14 articles · Updated · The Register · May 19
Google said it is increasing capital spending and emphasizing heavier token consumption to build always-on AI agents for search users.
Always-on agents require more persistent compute and inference capacity, tying the spending push directly to the broader boom in agentic AI development.
The company framed the outlays as infrastructure for next-generation search experiences rather than a short-term experiment, signaling deeper integration of AI into its core product.
The move underscores how the AI race is shifting from model launches to the costly deployment of continuously available services that demand sustained hardware investment.
As AI integration backfires for companies like Pizza Hut, what hidden operational risks are other firms ignoring?
While tech giants fund open-source security, is it enough to prevent the collapse of our digital supply chain?
With machine identities now outnumbering humans, how can we govern a world run by autonomous agents?
Google’s AI Surge: How Agentic AI, Gemini, and $20B Cloud Revenues Are Reshaping Search, Business, and Market Leadership in 2026
Overview
In May 2026, agentic AI systems moved from theory to real-world use, transforming how people search for information and how businesses operate. Driven by advancements like those revealed at Google I/O 2026, tools such as Google's Antigravity evolved from coding assistants into powerful features within Search, enabling users to create personalized, interactive experiences. These new AI agents help users manage complex tasks and long-term plans directly through Search, marking a major shift in digital interaction. This practical deployment of agentic AI signals a new era where intelligent agents reshape both everyday workflows and business processes.