Technology 15 sources · today Latest coverage 8 Oct 2026, 10:39 pm UTC

Google Cloud introduces Gemini agent to handle work across business apps

Google is pitching a persistent workplace agent that can execute tasks across business systems, as competition for corporate AI customers intensifies and initial access remains limited.

By Hushread Stories, written with AI from 15 outlets · First published 8 Oct 2026

In brief

  1. Google Cloud announced its Gemini agent for enterprise customers at the Gemini at Work 2026 event.
  2. The agent can research information, create documents, write code and coordinate other agents on workplace tasks.
  3. Each coworker agent gets its own Workspace account, email address, calendar and entry in the company directory.
  4. Google says the agent can handle tasks lasting hours or days and work across apps and devices.
  5. Initial access is described as a closed subscription preview, rather than a broadly available workplace product.
Google Cloud introduces Gemini agent to handle work across business apps
Source: 9to5Google

Timeline · 4 moments

4 moments Open the full timeline →

Dedicated workplace accounts enable persistent background agent work

The Next Web ↗

Task delegation and multiple model support are detailed

TechCrunch ↗

Google describes workflow automation connected to business systems

The Keyword ↗

Google Cloud announces Gemini agent at workplace event

9to5Google ↗

How it started

Google Cloud announced the Gemini agent at Gemini at Work 2026 on October 8, presenting it as a "universal agent for work" for enterprise customers, according to 9to5Google.

The launch puts Google into closer competition with OpenAI, Anthropic PBC and Microsoft Corp. for corporate customers, Bloomberg Technology reported. Its pitch is an assistant that carries out work across business systems, rather than only answering questions.

How it unfolded

At the October 8 announcement, Google Cloud CEO Thomas Kurian described Gemini as "your new single, universal agent for work," according to The Register. Google's own account in The Keyword said the agent connects to business systems to automate workflows from one prompt.

The capabilities described in the launch coverage include researching information, creating documents, writing code and coordinating other AI agents. Technology reported that the agent can complete tasks lasting hours or days.

The system can also plan and execute tasks, delegate work to subagents and use multiple AI models, according to TechCrunch. SiliconANGLE reported that it can act autonomously across web, mobile and desktop devices.

A separate workplace identity is part of the design. The Next Web reported that each coworker agent gets its own Workspace account, email address, calendar and company directory entry, and can continue working for days after a laptop closes. The launch coverage also describes dedicated Gmail, Calendar and Drive storage, according to Technology.

Where it stands

The agent has been announced for enterprise customers, but that does not establish broad availability. AI on vc.ru describes access as a subscription offering within a closed preview. The Verge reports that the tool will be available within Gemini Enterprise.

The reported features remain launch claims, not evidence of performance in everyday workplaces. CBS attributes the ability to write and run code, integrate with apps such as Gmail and help with routine work to Google.

What to watch

The immediate question is when access will extend beyond the closed preview described by AI on vc.ru. The supplied coverage gives no date for a broader rollout.

For businesses testing it, the practical issue is whether it can reliably complete the long-running, cross-system tasks being advertised. Its ability to delegate work to other agents, reported by TechCrunch, makes that a more demanding test than generating a single answer or document.

Spotted an error? Email [email protected] and we'll fix it.

More in Technology

All →