Google Cloud Next '26 Opening Keynote

Google Cloud Next '26 took place April 22-24, 2026, in Las Vegas, with the Opening Keynote (titled "The Agentic Cloud") delivered primarily by Google Cloud CEO Thomas Kurian. It centered on the shift from AI experimentation/pilot projects to large-scale production deployment of AI agents in enterprises. Core Theme: The Blueprint for the Agentic Enterprise The keynote presented a unified stack to help organizations move beyond chatbots and simple AI adoption toward autonomous, goal-oriented AI agents that act on data, integrate with business processes, and drive transformation at scale. Key idea: Enterprises need a complete system connecting chips, models grounded in enterprise data, agents/applications, and secure infrastructure. Major sections included: Agentic Enterprise Blueprint (early in the talk). Gemini Enterprise Agent Platform — the flagship announcement. AI Hypercomputer (infrastructure layer). Agentic Data Cloud. Key Announcements Gemini Enterprise Agent Platform A comprehensive platform for building, scaling, governing, securing, and optimizing production-grade AI agents. It includes tools like Agent Studio, runtime environments, identity/gateway controls, a registry, evaluation/observability features, and strong security. It serves as the "connective tissue" between data, people, and business goals. Gemini Enterprise App Positioned as the primary daily work environment for businesses in the agentic era. It features an Agent Designer, an Inbox for managing agent activity, support for long-running agents, Skills, Projects, and deeper integration into workflows (including Workspace-related capabilities). AI Infrastructure Advances Introduction of eighth-generation TPUs (Tensor Processing Units) for more powerful and efficient handling of demanding AI workloads. Enhancements to storage and networking to support agent-scale operations. The AI Hypercomputer as the foundational infrastructure layer optimized for AI models and agents. Agentic Data Cloud Updates to make data architectures suitable for agent-scale operations (moving beyond human-speed legacy systems) with better grounding, actionability, and integration. Additional emphasis was placed on security (using AI to defend against threats, secure agent deployment, and provide end-to-end visibility/control) and real-world customer examples demonstrating tangible impact.