Google Cloud introduces Gemini agent for workplace tasks

Google Cloud has introduced Gemini agent, an artificial intelligence system designed to carry out multistep workplace tasks for individuals, teams and specific organizational roles. Google says users can give it an objective and have it plan the work, use tools and return completed materials, including documents, email and code. Announced at the Gemini at Work 2026 event on October 8, the agent is intended to work across Google Workspace and third-party services. Google says it can be accessed through services including Microsoft 365 and Slack, and can connect with business tools and systems. The company describes the product as able to answer questions, create content and media, and write or run code. The agent can use subagents to divide complex assignments into smaller tasks and coordinate the steps, Google says. Because it runs in the cloud, work can continue for hours or days after an employee closes a laptop, according to the company. The system is also designed to operate through different devices and channels, including web and mobile access.