Google Cloud CEO Thomas Kurian used the Cloud Next 2026 keynote in Las Vegas to formally declare the “agentic enterprise” as the company’s defining strategic bet. The centerpiece: the Gemini Enterprise Agent Platform, a production environment for building, deploying, governing, and scaling autonomous AI agents across business operations.

From Intelligence to Action

The platform represents Google Cloud’s shift from treating AI as an analytics layer to positioning it as operational infrastructure. Karthik Narain, Google Cloud’s chief product and business officer, described the vision in an interview with SiliconANGLE: “The vision of the agentic enterprise is that we build a system that creates a connective tissue between people, data and the business outcomes of the enterprise.”

The stack has five layers, according to the Google Cloud Community blog: the AI Hypercomputer (custom silicon), Agentic Data Cloud (enterprise context for agents), Agentic Defense (autonomous security), the Gemini Enterprise Agent Platform itself, and Agentic Task Force (pre-built agents for specific workflows).

Kurian framed the platform as the evolution beyond Vertex AI. “Gemini Enterprise is now the end-to-end system for the agentic era. The connective tissue between your data, your people, and your goals,” he said during the keynote, according to the Google Cloud Community recap.

Governance Tools for Agent Fleets

The platform introduces three governance primitives aimed at enterprise IT teams, as Computerworld reported.

Agent Identity assigns a unique cryptographic ID to each agent, creating an auditable trail for every action mapped to authorization policies. Agent Registry acts as a central library indexing all internal agents and skills, ensuring only approved assets are accessible. Agent Gateway provides what Michael Gerstenhaber, VP of product management for Cloud AI, called “air traffic control” for agents, delivering secure, unified connectivity between agents and tools across environments.

Mike Leone, VP and principal analyst at Moor Insights & Strategy, told Computerworld that Google “wants every agent action to leave the same audit trail a payroll transaction does. That’s a high bar.”

New Capabilities in Gemini Enterprise

Alongside the platform, Google shipped several features in the Gemini Enterprise app itself. Long-running agents can now operate autonomously in cloud sandboxes, handling multi-step workflows like financial reconciliations that span days. An Inbox feature provides a centralized command location to monitor agent activity, receive notifications, and track task completion, as Computerworld detailed.

Agent Designer, a no-code builder for schedule- or trigger-based agents, is now generally available. An Agent Gallery provides access to third-party agents from Adobe, Lovable, and ServiceNow. Gemini Enterprise costs $30 per user per month for large organizations, according to Computerworld.

Scale Numbers

Google provided internal usage metrics to validate the stack. First-party models now process over 16 billion tokens per minute via direct API use, a nearly 60% increase from the previous quarter, according to the Google Cloud Community blog. Gemini Enterprise users grew 40% quarter-over-quarter. And during the keynote, Alphabet CEO Sundar Pichai disclosed that 75% of all new code at Google is now AI-generated and approved by engineers.

The Hyperscaler Alignment

Google’s announcement lands within 48 hours of similar agentic enterprise pushes from AWS and Microsoft. AWS expanded its Bedrock Managed Agents offering earlier this week. Microsoft reported a $37 billion AI revenue run rate in its Q1 2026 earnings, with CEO Satya Nadella citing the “agentic computing era” as the next frontier.

Ed Anderson, research VP at Gartner, told Computerworld: “Implementing agentic processes is seen by most vendors as the next big opportunity to drive adoption of AI technologies in enterprise environments. There will be a lot of competition in this area.”

Leone noted that Google’s edge is “full-stack ownership across infrastructure, models, data, security, and end-user apps,” but flagged a risk: “My bigger concern is that a typical IT buyer still struggles to figure out what to buy from Google and what each product actually includes as the portfolio and services have grown and evolved quickly.”

The message from all three hyperscalers is now identical: agents are not an experiment. They are business operations infrastructure, and the platform that governs them at scale wins the next enterprise cycle.