Google Cloud Unveils Universal Gemini Agent to Transform Enterprise Work

Google Cloud Introduces Universal AI Agent for Enterprise Work

Google Cloud CEO Thomas Kurian announced the Gemini agent at the Gemini at Work 2026 event on Thursday, positioning it as a universal AI agent for enterprise work that fundamentally changes how employees delegate and complete tasks1

1

3

3

. The autonomous AI assistant operates across web, iOS, Android, Windows, Mac, command line, Google Workspace, Microsoft 365, and Slack without requiring a dedicated interface4

4

. Currently available in private preview for enterprise customers, the Gemini agent represents Google's attempt to consolidate its AI offerings into a single platform capable of handling entire assignments rather than individual prompts2

2

.

The announcement arrives amid intensifying competition in enterprise AI. Microsoft introduced an updated Copilot on September 25, followed by OpenAI's always-on dots agents four days later. Anthropic announced native Claude integration with Google Docs, Sheets, and Slides on October 6, while SpaceXAI introduced persistent Grok Bot agents in August2

2

. Google says nearly 90% of Fortune 100 companies are Gemini Enterprise customers, with close to 500 Cloud customers processing more than 1 trillion tokens each over the past year3

3

5

5

.

Persistent Memory and Cross-Device Functionality

The Gemini agent maintains a single set of memories, context, and personalization across all communication channels and devices because it runs in the cloud1

1

4

4

. Employees can start an assignment on one device, leave it running, and resume from another without reconstructing context2

2

. The agent organizes knowledge across four distinct memory types: session memory tied to active tasks, semantic memory assembled from documents and conversations, procedural memory capturing workflows, and episodic memory recording prior completed assignments3

3

.

Inside Google Workspace, the agent works directly within Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar3

3

5

5

. The persistent execution capability allows the agent to manage workflows lasting multiple days, coordinate parallel assignments, and respond to scheduled tasks or events2

2

. Users can delegate objectives rather than provide step-by-step instructions, with the agent determining how to complete work and delivering finished results3

3

.

Coworker Agent Configuration with Dedicated Identity

The Gemini agent can function as a persistent digital coworker with responsibilities shared across teams2

2

. Managers can create specialized agents—such as an Event Planner Agent—that receive their own Workspace account, email address following the @agents.company.com format, calendar, Drive storage, and company-directory presence2

2

4

4

. Employees can assign work to these coworker agents much as they would to human colleagues, including by mentioning them in Google Chat or requesting document edits2

2

.

Each coworker agent is granted its own cryptographically attested identity and permissions managed through enterprise security systems, with actions recorded in audit logs2

2

3

3

. These agents have access only to the information that team members choose to make available, rather than inheriting broader organizational access3

3

5

5

. The agent can also assemble temporary groups of specialized sub-agents for complex work, distributing different responsibilities among them2

2

.

Multi-Model Orchestration and Cost Management

Rather than relying on a single model, the Gemini agent selects from multiple AI models depending on the task3

3

. Simple tasks might run on Gemini Flash, a small model for day-to-day work, while complex long-horizon assignments use flagship frontier models like Argon4

4

. The system currently supports Google's Gemini models and Anthropic Claude models, with support for additional proprietary and open-weight AI models planned2

2

3

3

.

A Smart Routing capability evaluates each workload and assigns it to whichever model balances quality and cost most favorably3

3

5

5

. Administrators can configure spending thresholds at the project level, and the agent will suspend operations automatically once those limits are hit3

3

4

4

. Google noted that per-token prices have dropped 98% since 2024, though enterprise AI volume has greatly increased4

4

5

5

.

Enterprise Integration and Reusable Skills

The Gemini agent connects securely with collaboration tools including Confluence, Microsoft Office, Teams, Slack, and Workspace, as well as development tools like Git and Jira4

4

5

5

. It also integrates with enterprise platforms including Salesforce and ServiceNow, and databases such as BigQuery, Databricks, Postgres, and Snowflake4

4

5

5

. Users can build reusable skills that act as instructions, knowledge, and workflows—essentially prompts that teach agents how to perform specific multistep tasks4

4

.

For data scientists and engineers, Google is extending the Gemini agent with machine learning skills and tools. Users can describe an outcome in plain language, and the agent can generate PySpark code, provide a notebook to edit and test it, train a model, and troubleshoot issues autonomously4

4

. For business users, the agent can use operational BigQuery reporting skills and the Knowledge Catalog to construct and save queries, allowing teams to run reports without incurring additional token costs4

4

.

Industry-Specific Capabilities and Customer Deployments

Google announced industry-specific versions currently in preview for financial services and legal teams, with government, healthcare, and retail versions coming3

3

5

5

. The financial services version draws on data from FactSet and LSEG and includes more than 50 skills, while the legal version integrates with document management platforms including NetDocuments and iManage3

3

. BNP Paribas has rolled out the platform to over 65,000 staff using LLM@CIB, an internal generative AI assistant supporting credit memo preparation and customer advisory tasks3

3

5

5

.

Japanese insurer SOMPO has deployed more than 10,000 custom agents for 34,000 employees, reducing model development time from one week to one day3

3

5

5

. Tata Steel deployed more than 300 specialized AI agents within nine months for manufacturing safety, predictive supply chains, and customer complaints, reducing complaint turnaround time by 50%5

5

. Bradesco uses the Gemini agent for fiscal, accounting, and contractual analysis, reducing document review time from one hour to five minutes and risk inconsistencies by 60%, while delivering more than 10% financial efficiencies5

5

. Google's TPU 8i delivers up to 80% better price-performance than the previous generation, supporting these workloads5

5

.