Google debuts enterprise Gemini agent for work across apps

Google is introducing a work-focused Gemini agent that can take instructions from a single interface and carry out tasks across apps and devices, according to The Verge. Announced at Google’s Gemini at Work event, the tool matters because it points to a more centralized version of workplace AI: less jumping between apps, more delegating tasks from one place.

What Google is adding to Gemini Enterprise

The new agent is described by Google as a “universal” Gemini AI agent for the workplace. It will live inside the Gemini Enterprise app, where users can chat with it and assign work from one interface.

Google Gemini is already positioned as Google’s AI assistant, but this announcement pushes it further into enterprise task handling. Instead of only using Gemini inside individual Workspace products, Google says the agent can work in the background across apps and devices while keeping context in the cloud.

The Verge reports that the agent can operate directly inside Workspace apps such as Gmail, Drive, Docs, Sheets and Calendar. Google also says users can interact with it from mobile devices, desktop computers, the web and third-party apps including Slack and Microsoft 365.

That cross-surface design is the main point for teams. If it works as described, a worker could start from a chat-style request rather than opening each app separately to move a task forward. Google has not shared pricing, a broad release date or performance benchmarks in the source report.

The agent model: sub-agents and a coworker identity

Google says Gemini can use job-specific sub-agents to complete work. The company also describes a “coworker agent” mode with a dedicated identity and its own email address using an “@agents.company.com” format.

The practical implication is that Google is framing the agent less like a simple chatbot and more like a workplace participant that can be assigned responsibility. For operations, marketing, sales or administrative teams, that could make AI delegation easier to track, especially if the agent has a persistent identity.

Google also says Gemini will automatically choose an AI model suited to the task. The source does not specify which models are used for which tasks, so buyers should treat that as a high-level capability rather than a detailed deployment guide.

For small businesses and freelancers, the most relevant question is not whether the interface sounds convenient, but how much control and visibility users will have when the agent acts across sensitive work apps. The announcement, as reported, focuses on what the agent can connect to and how users interact with it; it does not provide details on admin settings, audit trails or approval flows.

Who gets access now

The new Gemini agent is not broadly available. The Verge reports that it is currently limited to enterprise customers in private preview.

That means most creators, freelancers and smaller teams should view this as a signal of where Google is taking Gemini rather than a tool they can necessarily adopt today. It is especially relevant for organizations already invested in Google Workspace and evaluating whether AI assistants should sit inside productivity suites or outside them.

Key access details from the report:

• The agent is part of the Gemini Enterprise app.
• It is available to enterprise customers in private preview.
• Google has not provided general availability details in the source report.
• The agent can be reached across mobile, desktop, web and some third-party apps, according to Google.

Google previously launched Gemini Spark in May for consumers, according to The Verge. That agent lets subscribers perform tasks across Workspace and the web, but it uses a separate interface inside the Gemini app. The enterprise agent appears to bring a related idea into a work-centered environment with broader enterprise positioning.

How to compare it with other AI work assistants

For readers comparing Gemini with ChatGPT or Claude, the useful comparison is not model quality; Google’s announcement, as reported, does not include benchmark claims. The clearer distinction is workflow location: Google is emphasizing a single enterprise interface tied to Workspace, connected devices and selected third-party apps.

That makes this announcement most important for teams trying to reduce tool-switching inside everyday work. If your company already lives in Gmail, Docs, Sheets and Calendar, an agent embedded in that environment may be easier to evaluate than a standalone assistant. If your workflows are spread across many non-Google systems, the third-party app story will matter more.

Emonarc readers can also place this alongside broader workplace AI moves, including Microsoft’s recent agent-focused Windows updates and our ChatGPT vs Claude vs Google Gemini guide. The near-term takeaway is simple: Google is making Gemini Enterprise a hub for assigning work, but access remains limited and many deployment details are still not public.