Google Introduces Agentic AI Capabilities to Gemini, Starting with Enterprise Deployment

From Conversational Assistant to Autonomous Business Agent

At a Google Cloud event on Thursday, the company unveiled a unified agent built on Gemini that moves beyond answering questions to executing tasks on behalf of users through a single interface. The announcement reflects a broader industry shift where AI tools are evolving from purely conversational experiences to systems that can take ownership of assigned work, generate code, schedule meetings, book travel, and more.

Google’s timing aligns with the emergence of consumer‑facing agents such as Meta’s Muse and messaging‑based assistants like Instinct, as well as the recent rollout of ChatGPT’s Dots. The company believes its position is strong, noting that Gemini already serves over 1 billion monthly active users and that nearly 90 percent of Fortune 100 firms now employ Gemini Enterprise in their operations.

CEO Sundar Pichai told the audience that focusing first on businesses will let Google tackle the “harder problems around security, scale, and performance” inherent in deploying powerful autonomous agents. Google Cloud CEO Thomas Kurian explained that the new agent accepts “objectives, not just instructions,” enabling it to plan work, employ custom skills, and connect to internal enterprise systems to meet its goals.

The platform automatically selects the most suitable model for a task, but users can override this choice, including by picking third‑party models—initially Anthropic’s Claude—while Google plans to broaden the model picker to include open‑source and private options later.

Requests can include attachments such as files, folders, or project bundles designed for specific workflows. The agent can integrate with a wide range of business tools, including Google Workspace, Microsoft 365, Slack, Jira, Confluence, Git, BigQuery, Databricks, PostgreSQL, Snowflake, and others. It also works securely with any Model Context Protocol (MCP) server, whether inside or outside the corporate network.

Users monitor the agent’s activity through a “tasks inbox” that displays its reasoning, delegation to sub‑agents, loading of specialized skills, code execution, and progress. The AI operates as a distinct employee within Workspace, complete with its own email address and context. It understands organizational structure, time zones, approval hierarchies, and calendar availability. Users can summon the agent by tagging, emailing, sharing, or adding it to group chats. Each action generates an audit trail attributed to the agent rather than a person.

Access will be available on iOS and Android mobile devices, Windows and Mac desktops, the command line, and within Google Workspace, Microsoft 365, ServiceNow, and Slack.

Early adopters include sportswear brand On, e‑commerce platform Shopify, and payment provider PayPal. Notable Gemini Enterprise customers span multiple industries, such as BNP Paribas, Bradesco, Merck, Orange Spain, Santee Cooper, SOMPO, Ulta Beauty, and Wesfarmers.

Google also introduced new flexible spending options—multi‑model orchestration, smart routing, and real‑time spend caps—to help enterprises better control their AI costs.