Google Unveils 1 Universal Gemini Agent to Orchestrate Enterprise AI Across Apps

3 articles · Updated · Computerworld · Oct 9

Google said the new Gemini agent is designed to run enterprise work from a single prompt window, acting as a personal assistant or role-based coworker with its own email, storage and persistent identity.
The cloud-based system can operate across Google Workspace, Microsoft 365, Slack, command lines, third-party apps and devices, while spawning sub-agents to handle multi-step workflows that can run for hours or days.
Google is also separating the agent from any single model: Gemini can already choose between Gemini and Claude for different tasks, with more private and open models planned to improve quality and cut cost.
Analysts said that model choice, role-specific agents and tighter identity, audit and permission controls are Google's main differentiators in a market where OpenAI, Microsoft, Anthropic and Meta already offer multiagent capabilities.
The broader test is whether CIOs and CISOs will trust autonomous agents with mission-critical processes, making governance, human oversight and clear decision limits central to enterprise adoption.