Google Introduces Gemini Agent for Workplace Tasks

Google Cloud has introduced a workplace artificial intelligence agent that, according to the company, can plan and carry out multistep assignments across business applications, then return finished work such as documents, email and code. The Gemini agent was announced October 8 at the Gemini at Work 2026 event, with businesses as its initial audience. Google says users can give the agent an objective rather than specify each step. It can use tools and company systems, divide complex tasks among temporary subagents and coordinate their work. Because it runs in the cloud, Google says assignments can continue for hours or days after an employee closes a laptop. The agent is designed to work across Google Workspace and third-party services, including Microsoft 365 and Slack. Google also says it can connect to business tools and databases, and can be accessed through different devices and channels. The company says its system currently routes tasks between Google’s Gemini models and Anthropic’s Claude models, with more model options planned. Organizations can also create what Google calls coworker agents, assigning them roles such as project manager or analyst.