At a Google Cloud event held on Thursday, the tech giant unveiled its plans to push Gemini AI into what it calls the "agentic age." The centerpiece is a new unified agent capable of both answering questions and actually completing tasks on a user's behalf—all through a single interface.

This development reflects a broader industry shift, as AI tools evolve from simple conversational assistants into systems that can take charge of assigned work: writing code, scheduling meetings, booking appointments and travel, and much more. It also arrives amid a wave of consumer-facing agents, including Meta's Muse, messaging-based tools like Instinct, and the recently launched Dots from ChatGPT.

Google appears well-positioned to scale its agentic ambitions. During the event's opening, CEO Sundar Pichai revealed that Gemini now boasts over 1 billion monthly active users. He also pointed out that nearly 90% of Fortune 100 companies have adopted Gemini Enterprise in their workplaces. Because of this strong corporate traction, Google will roll out the agent to businesses first, with a consumer release to follow later.

Pichai explained that this business-first approach lets the company tackle the "harder problems around security, scale, and performance" that come with deploying such powerful agents. Thomas Kurian, CEO of Google Cloud, described the new agent as something that can be given "objectives, not just instructions." In practice, that means it can plan out work, leverage custom skills and tools, and plug into a company's internal systems to reach its goals.

By default, the AI selects the most suitable model for a given task, though users can step in and pick a model themselves—including third-party options, beginning with Anthropic's Claude models. Google indicated that the model picker will eventually expand to cover open source and other private models. Requests can carry attachments such as files, folders, or other project materials tailored to specific workstreams, including combinations of files and skills.

The agent integrates with a business's data and systems, spanning Google Workspace, Microsoft 365, Slack, Jira, Confluence, Git, BigQuery, Databricks, Postgres, Snowflake, and more. It can also securely connect and operate with any Model Context Protocol (MCP) server, whether inside or outside the company's network.

A "tasks inbox" interface lets users monitor what Gemini is up to—observing its reasoning, how it delegates work to subagents, which special skills it loads, its code, and its overall progress. Notably, the AI gets its own Workspace account, effectively treating it like just another coworker. That means it has a dedicated email address and its own context. It understands who belongs to which team, their time zones, who holds approval authority, what's scheduled on calendars, and similar details. Users can reach the agent by tagging it, emailing it, sharing with it, or adding it to a group chat. As it operates, it generates an audit trail credited to the agent rather than a person. Access is available across iOS and Android mobile devices, Windows and Mac desktops, the command line interface, Google Workspace, Microsoft 365, ServiceNow, and Slack.

Early testers of the agent included sportswear brand On, Shopify, and PayPal. Among Gemini Enterprise customers are major businesses such as BNP Paribas, Bradesco, Merck, Orange Spain, Santee Cooper, SOMPO, Ulta Beauty, and Wesfarmers. Google also announced new flexible spending options—featuring multi-model orchestration, smart routing, and real-time spend caps—designed to give businesses tighter control over their enterprise AI costs.