Google Cloud's Agentic Push

Google Cloud used its April 22 Cloud Next keynote to put AI agents at the center of its enterprise pitch, not just its model lineup. (blog.google) At the Las Vegas event, Google introduced a new Gemini Enterprise Agent Platform and said the “agentic enterprise” is now its roadmap for customers. Google also said nearly 75% of Google Cloud customers use its artificial intelligence products, and 330 customers processed more than 1 trillion tokens each over the past 12 months. (blog.google) Chief Executive Sundar Pichai said Google’s first-party models now process more than 16 billion tokens per minute through direct customer application programming interface use, up from 10 billion last quarter. He said just over half of Google’s machine learning compute investment in 2026 is expected to go to the Cloud business. (blog.google) An AI agent is software that can take a goal, call tools, and complete multi-step work with limited human input. Google’s sales pitch is shifting from single chatbots to systems that can route work among multiple agents and connect them to company data, software, and security controls. (docs.cloud.google.com, cloud.google.com) Google’s product stack reflects that shift. Vertex AI Agent Builder now emphasizes multi-agent workflows, while Agent Development Kit is positioned as the framework for orchestrating how agents “think, reason, and collaborate” under deterministic guardrails. (cloud.google.com, docs.cloud.google.com) Google is also pushing interoperability, not just its own models. Its Agent2Agent protocol is described as a way for agents built with different frameworks and vendors to communicate, and Google says more than 50 partners including Salesforce, ServiceNow, Box, Deloitte, Elastic, UiPath, and UKG are involved. (cloud.google.com) The company paired the product launch with a customer showcase. In a roundup published April 22, Google highlighted Capcom, Citi Wealth, Home Depot, Mars, and other users as examples of agents moving from pilots into production tasks such as game testing, customer service, and research. (blog.google) Google is also underwriting the ecosystem around the platform. It announced a $750 million partner fund for agent development and deployment, plus new forward-deployed engineering support and an Agent Gallery inside Gemini Enterprise for partner-built agents. (cloud.google.com) Reuters reported April 22 that Google is making AI agents “a linchpin” of its plan to monetize artificial intelligence through enterprise software, as OpenAI and Anthropic also push harder into business customers. Reuters said Thomas Kurian told attendees at the conference that “the experimental phase is behind us.” (usnews.com) That changes what enterprise buyers will compare when they choose a cloud vendor. The questions now extend beyond model quality to how agents are governed, how they connect to existing systems, how their outputs are evaluated, and what service levels a provider will stand behind in production. (docs.cloud.google.com, cloud.google.com) Google’s message in Las Vegas was that the next cloud sale is not just compute plus a model. It is a managed system for building, coordinating, and policing fleets of agents inside large companies. (blog.google, usnews.com)