Google Cloud is introducing a significant update to Gemini Enterprise, aiming to make AI agents a core component of enterprise automation. The company envisions AI moving beyond simple tasks to handle complex, multi-step processes currently done by humans.
Announced at Google Cloud Next 2026, the new Gemini Enterprise is designed to be a "secure, collaborative autonomous engine." This shift addresses a key challenge: moving AI out of isolated productivity roles into a governed system where agents have distinct identities for monitoring and management.
Many businesses have explored AI but face a "human-in-the-loop" bottleneck, requiring constant human prompting for multi-stage projects. Gemini Enterprise seeks to overcome this by enabling "long-running" AI agents managed through the new Gemini Enterprise Agent Platform. It offers low-code and no-code tools like Agent Designer for users to build their own agents, combining generative intelligence with strict business rules to ensure compliance.
Users can codify expertise into reusable "Skills" for specific workflows, ensuring consistency and avoiding repeated context explanations. A new "Inbox" provides a centralized hub for monitoring agent activity, categorizing notifications by action needed, errors, or completion status.
Features like Projects and Canvas foster collaboration, allowing humans and agents to co-create in shared workspaces. Projects integrate with Google Workspace and Microsoft OneDrive, while Canvas enables collaborative work on documents and slides with AI agents.
Gemini Enterprise now supports Bring Your Own Model Context Protocol, allowing agents to connect to internal tools and servers. An Agent Marketplace will feature specialized third-party agents from partners like ServiceNow and Oracle.
Google is emphasizing governance with "Agent Identity," a traceable digital ID for each agent to enable tracking and auditing. This facilitates "least privilege access" and helps mitigate risks like data leaks and prompt injection attacks through an Agent Gateway.
The future workplace, as envisioned by Google, will be a hybrid of human experts and persistent AI agents. These updates aim to provide businesses with a comprehensive understanding of their AI agents' actions and motivations, backed by robust controls.