Google Cloud announces a universal Gemini agent for work
Google says the enterprise agent can plan tasks, use company tools and context, and return completed work inside documents, inboxes, and developer environments.
One agent across business tasks
Google Cloud announced the Gemini agent at its Gemini at Work 2026 event on October 8. The company describes it as a universal agent for enterprise work, combining questions and answers, content creation, and coding in one prompt interface. Google says the agent can plan a job, use skills and tools, connect to a customer’s business systems, and return work inside documents, email, or developer environments. This is a broader product announcement than a new standalone model: the promised value is coordinating work across systems using the organization’s context.
Controls and availability remain to be tested
Google says the agent can select a model for each task and includes cost controls as well as security, administration, and governance features for enterprise customers. Those are the company’s descriptions, not independent measurements of task success, cost savings, or protection against unauthorized actions. The short public announcement does not specify a general availability date, pricing, or which capabilities every customer can use immediately. Businesses evaluating it would still need to test the relevant connectors, permissions, and review controls in their own environment before relying on the agent for consequential workflows.