Google Cloud, the cloud computing division of Alphabet, introduced the Gemini agent on October 8, 2026, offering businesses a single AI agent that can answer questions, complete tasks, create content and write code. The launch comes as technology companies compete to develop AI systems capable of carrying out work autonomously rather than simply responding to user prompts.
Announced at the Gemini at Work 2026 event, the new agent is designed to help employees delegate tasks through a single interface. According to Google Cloud, users can provide an objective, and the agent can plan the steps, use tools, connect to business systems and return completed work in documents, email and developer environments.
The Gemini agent works within Google Workspace applications, including Gmail, Docs, Sheets and Calendar. It can also be accessed through Microsoft 365 and Slack, allowing businesses to use the technology across their existing work environments.
AI agent can complete tasks across business systems
Google said the Gemini agent can support a range of workplace activities, from answering questions and creating content to writing and running code. Instead of requiring users to provide instructions for every step, the system is designed to determine the work required to achieve a given objective.
The agent can select the AI model best suited to each task. At launch, it uses Google’s Gemini models and models from Anthropic, including Claude. Google said it plans to add support for other models in the future.
The platform also includes cost controls, security measures, administration features and governance capabilities designed for enterprise customers. These functions are intended to help organisations manage how AI agents access information and perform tasks within business environments.
Businesses can create AI coworker agents
Another feature allows organisations to create what Google calls “coworker agents”. These agents can function as team members, taking on defined responsibilities and working across tasks assigned by employees.
Each coworker agent can have its own email address and access only the information made available to it. This allows businesses to assign work to an AI agent while maintaining controls over the information it can access.
Google said the Gemini agent can work across business systems and return completed outputs within the applications employees already use. This approach is intended to reduce the need to move between separate AI tools and workplace applications.
Versions designed for financial services and legal work are available in preview. Google said versions for government, healthcare and retail are planned for release in the future.
Google enters a growing market for autonomous AI
The launch comes amid increasing competition among technology companies developing AI agents that can perform multi-step tasks on behalf of users.
OpenAI introduced its always-on agents, called dots, in September 2026. These agents are designed to pursue user goals across applications. Meta also released Muse earlier in September, a personal AI agent capable of activities such as shopping, booking travel, sending emails and making payments.
Google’s announcement expands its efforts to bring AI into everyday business workflows. By combining task planning, access to company systems, model selection and workplace applications, the Gemini agent is designed to help organisations automate work across multiple functions.
The shift towards autonomous AI also introduces new requirements around security, permissions and oversight. Business agents may need access to emails, documents, calendars and internal systems to complete tasks, making it important for companies to control what information each agent can use and which actions it can perform.
Google said its Gemini agent has the security, administration and governance features required by enterprise customers. The company is positioning the product as a single interface through which employees can delegate work and receive completed outputs.
The launch highlights the broader transition in enterprise AI from conversational tools towards systems that can plan and execute tasks. As Google Cloud and its competitors expand their offerings, the ability to integrate AI agents with existing business applications and manage them securely is likely to remain central to adoption.
