Google Launched the Universal Gemini Work Agent: The Digital Coworker With Its Own Email | inicIA

Direct answer

On October 8, 2026, at the Gemini at Work 2026 conference, Google Cloud CEO Thomas Kurian introduced the universal Gemini agent: a single AI agent for work that answers questions, handles knowledge work, creates images and media, and writes and runs code — all from one prompt window. The difference from a traditional assistant: it takes objectives, not instructions. It plans the work, picks tools and skills, connects to company systems, and returns finished work inside Gmail, Drive, Docs, Sheets, Chat, Calendar, Microsoft 365, and Slack. It runs persistently in the cloud: a job lasting hours or days keeps going even after you close your laptop.

What it includes, in facts

Governance: the real announcement

Google shipped the agent with enterprise-grade controls: cryptographically attested agent identity, OAuth-propagated permissions, per-agent audit trails (every action recorded under the agent's name, not a person's), an Agent Sandbox for isolated execution, and Agent Gateway, an "AI firewall" enforcing corporate policy in real time. On costs: smart routing across models and real-time per-project spend caps that pause the agent automatically. On data: Knowledge Catalog (business definitions), Smart Storage, and a Borderless Lakehouse querying Amazon S3 and Azure Data Lake with no variable egress fees.

What it means for your business

  1. The agent becomes a team member with credentials: that demands the same rigor as a new hire — minimum access, permission reviews, audit — from day one.
  2. Objectives replace prompts: work starts by defining the expected outcome and the rules, not by writing step-by-step instructions.
  3. Cost is controlled per project: real-time spend caps are the practice you should demand of any agentic implementation, built or bought.
  4. Businesses first, consumers later: Google said the initial focus is business. If your company already lives in Workspace or Microsoft 365, this is the lane automation will arrive through.

Early testers include On, Shopify, and PayPal; Gemini Enterprise customers already include BNP Paribas, Bradesco, Merck, Orange Spain, Santee Cooper, SOMPO, Ulta Beauty, and Wesfarmers. Vertical specializations are in preview for financial services and legal (government, healthcare, and retail come next).

Related links

Frequently asked questions

What is the universal Gemini agent?

A single AI agent for work unveiled by Google Cloud on October 8, 2026: it takes objectives, plans, uses company tools and systems, and returns finished work inside the apps the team already uses. It runs persistently in the cloud for hours or days.

Can the agent have its own email?

Yes. You can create a "coworker agent" with its own Workspace account: email, calendar, Drive, and a seat in the corporate directory. Its actions land in an audit trail under its name.

Which models does it use?

It picks between Google's Gemini models and Anthropic's Claude models per task, with other open and private models planned. Smart routing balances performance and cost.

What security controls does it include?

Attested agent identity, OAuth-propagated permissions, per-action audit, execution sandbox, a gateway enforcing policies in real time, and per-project spend caps that pause the agent automatically.

Take the next step

If you want an agent that works toward clear objectives inside your systems — with identity, permissions, and audit from day one — talk to inicIA on WhatsApp: https://wa.me/17866120930

Chat on WhatsApp

By inicIA