At a Google Cloud event on Thursday, the company announced it is bringing Gemini AI into the agent era, launching a unified agent that can not only answer questions but do everything for you from a single interface.
The move comes as AI tools move beyond mere conversational experiences to tools that can take ownership of assigned tasks, generate code, schedule meetings, make reservations, travel, and more. This follows the rise of consumer-facing agents like Meta’s Muse, agents that work via messaging like Instinct, and even the recent debut of ChatGPT’s Dots.
The company still has enough prospects to achieve the scale of its agent efforts. As Google CEO Sundar Pichai pointed out at the beginning of the event, Gemini currently has more than 1 billion monthly active users. He also noted that nearly 90% of Fortune 100 companies currently use Gemini Enterprise in their work.
Given Gemini’s adoption in enterprises, Google plans to focus on bringing this agent to enterprises first, before rolling it out to consumers.
According to Pichai, this will allow the company to solve “the tougher questions around security, scale, and performance” that come with launching such a powerful agent.
Google Cloud CEO Thomas Kurian explained that new agents can be given “goals, not just instructions.” This means you can plan your work, use custom skills and tools, and connect to your company’s internal systems to achieve your goals.
By default, the AI chooses the best model to complete the task, but you can also take over the model selection. This also includes third-party models such as Anthropic’s Claude model. Google said it plans to expand the model picker to include open source models and other private models in the future.
Requests can include attachments such as files, folders, or other projects designed for a particular workstream, such as a combination of files and skills. Agents can connect to your business’s data and systems, including Google Workspace, Microsoft 365, Slack, Jira, Confluence, Git, BigQuery, Databricks, Postgres, and Snowflake.
It can also securely connect to and work with Model Context Protocol (MCP) servers inside and outside the corporate network.
Users can track what Gemini is doing through the “Task Inbox” interface, where they can see Gemini’s thought process, delegate tasks to subagents, load special skills, code, and progress.
In particular, the AI has its own workspace account, just as if it were another colleague. That means it has its own email address and its own context, Google said. Know who at your company is on which team, time zones, who needs to approve items, what’s on people’s calendars, and more. Users can call agents by tagging them, emailing them, sharing them, or adding them to a group chat. When you perform an action, an audit trail is written that is attributable to the agent, not a human.
The agent can be accessed from iOS and Android mobile devices, Windows and Mac desktops, a command line interface, Google Workspace, Microsoft 365, ServiceNow, and Slack.
The agent noted that early testers include sportswear brand On, Shopify, and PayPal, and that Gemini Enterprise’s customers include large companies such as BNP Paribas, Bradesco, Merck, Orange Spain, Santee Cooper, Sompo, Ulta Beauty, and Wesfarmers.
Google said new flexible spending options such as multi-model orchestration, smart routing, and real-time spending caps will help businesses better manage enterprise AI costs.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
