Google brings agentic AI to Gemini for businesses

Google announced a new Gemini agent at a Google Cloud event on Thursday, initially targeting businesses before a later consumer rollout. The company says the agent can take on objectives, plan work and carry out tasks through a single interface.

Users can let the agent choose an AI model or select one themselves. The initial third-party option includes Anthropic’s Claude models; Google says it plans to add open-source and other private models later. The agent can use custom skills and tools, connect to company systems and work with Model Context Protocol (MCP) servers inside or outside a company network.

A “tasks inbox” shows the agent’s progress, including task delegation to subagents. The agent also gets its own Google Workspace account and email address, with context about teams, time zones, approvals and calendars. Its actions are recorded in an audit trail attributed to the agent.

Google says Gemini has over 1 billion monthly active users and nearly 90% of Fortune 100 businesses use Gemini Enterprise. The company says it will start with businesses to address security, scale and performance challenges before bringing the agent to consumers. Early testers include sportswear brand On, Shopify and PayPal.