Google Cloud Announces Gemini Agent for Enterprise Workflow Automation
Google Cloud announced Gemini Agent, a general-purpose AI work agent designed to automate complex, multi-step enterprise tasks across Google Workspace and third-party services. By specifying a high-level goal, users can delegate tasks such as generating presentations, writing code, or acting as team roles directly to the agent. This marks a major shift from simple conversational AI assistants to autonomous enterprise agents capable of operating as functional team members. Deep integration into daily productivity tools allows organizations to streamline cross-departmental operations and reduce repetitive manual tasks. Gemini Agent is currently in private preview and notably supports both Gemini and Anthropic's Claude models, with plans to support additional open and closed-source models. Google will also provide APIs allowing developers to integrate Gemini Agent functionalities into third-party enterprise apps.
## BACKGROUND
Enterprise AI agents extend traditional rule-based workflow automation by leveraging large language models to reason, plan, and execute complex workflows autonomous of predefined scripts. Google Workspace Intelligence acts as the underlying contextual layer, connecting business data across Gmail, Drive, and Calendar to ground AI outputs in relevant corporate context.