Google Cloud launches Gemini Agent with autonomous coworker email accounts

Autonomous coworker agents for the enterprise

At the Gemini at Work event, Google Cloud introduced the Gemini Agent, a single AI that can take a goal, execute it across Gmail, Docs, Sheets, Slides, Drive, Chat and Calendar, and return a finished result. The agent can stay active in the cloud for hours or days, even after a user logs off. A standout feature is the “coworker agent”: a manager defines a role—such as events coordinator—and the system creates a dedicated Workspace account with its own email address, calendar, Drive storage and a listing in the company directory.

Deep integration and safety architecture

The agent connects to Salesforce, ServiceNow, Jira, Snowflake, BigQuery and any server that speaks Model Context Protocol, an open-standard plug for AI-software interaction. It can orchestrate both Google’s Gemini models and Anthropic’s Claude models, routing jobs to the most suitable model for quality and cost. To address safety, each agent receives a cryptographically attested digital ID, and all actions are recorded in an audit trail attributed to the agent rather than a human. Execution occurs inside an Agent Sandbox, with traffic filtered through an “AI network firewall” called Agent Gateway.

Pricing, preview status and rollout plan

Google disclosed no pricing or general-availability date. The financial services and legal versions are currently in preview, while government, healthcare and retail are listed as “coming soon.” The company noted that nearly 500 customers each processed more than one trillion tokens over the past year and introduced real-time spend caps that pause an agent when a project hits its limit.

The announcement marks a significant step toward autonomous AI workflows, but the absence of concrete cost and timeline details leaves enterprises awaiting further clarification.