Google's enterprise agent push

Google used Cloud Next on April 22 to roll out Gemini Enterprise Agent Platform, a new system for companies building and running large groups of artificial intelligence agents. (blog.google) An artificial intelligence agent is software that can take a goal, use tools, and complete a sequence of steps instead of answering a single prompt. Google said its new platform combines Vertex AI model services with agent integration, security, DevOps, and management tools in one product. (docs.cloud.google.com, blog.google) Google tied the launch to Vertex AI Agent Engine, which the company describes as a managed service for deploying, scaling, observing, and governing agents in production. The documentation lists runtime, sessions, Memory Bank, code execution, observability, and governance controls including threat detection and Identity and Access Management-based agent identity. (docs.cloud.google.com) Google also pushed Agent2Agent, or A2A, an open protocol for letting one agent discover and work with another agent across different systems. In Google’s documentation, an A2A agent publishes an “AgentCard,” which works like a business card describing what the agent can do before another agent sends it work. (docs.cloud.google.com) The company framed the launch as part of a broader “agentic enterprise” push at Cloud Next in Las Vegas. Thomas Kurian said in his keynote transcript that Google now offers a “vertically optimized stack” instead of a patchwork of separate systems, running from models and chips to enterprise software. (cloud.google.com) Google used customer-adoption numbers to support that pitch. 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 last 12 months, and 35 customers reached 10 trillion tokens. (blog.google, cloud.google.com) Google also pointed to named adopters to show agents moving from pilots into operations. In a Cloud Next customer roundup published April 22, Google cited companies including Capcom, Citi Wealth, Home Depot, and Mars as users of its agent systems for tasks such as game testing, customer service, and research workflows. (blog.google) The competitive angle is straightforward: companies want agents that can remember context, call tools, and work across departments without forcing developers to stitch together separate products. Google’s answer is to package models, infrastructure, data, security, and interoperability into one enterprise stack. (blog.google, cloud.google.com) The immediate test is whether enterprises buy the full stack or keep mixing vendors. Google’s message at Next was that the agent race is shifting from chatbots to production systems that companies can monitor, secure, and run at scale. (blog.google, docs.cloud.google.com)