Google Cloud unveils persistent Gemini Agents for enterprise tasks

Git: Git is the distributed version control system used for tracking changes in code repositories. The announcement includes Git among the development tools the Gemini agent supports. Agents can therefore interact with codebases as part of research, coding, and documentation tasks. Jira: Jira is Atlassian's project management and issue tracking software widely used in software development. It is included among the systems the new Gemini agent can connect to for coordinating work. This supports agents managing complex, multi-day assignments involving development workflows. Claude: Claude refers to Anthropic's large language models used for reasoning and task execution. The Gemini agent can already leverage Claude models for specific workloads alongside Google's own Gemini models. Support for additional models from other providers is planned to give enterprises flexibility. Google: Google develops and operates the Google Cloud platform and its Gemini family of AI models. In this news, the company announced a new universal persistent Gemini agent designed for long-running enterprise tasks at its Gemini at Work 2026 event. The agent integrates across Google Workspace, Microsoft 365, and Slack while supporting multiple underlying models including those from Anthropic. BigQuery: BigQuery is Google's serverless data warehouse for large-scale analytics. It is listed among the data platforms the new agent can connect to for analysis and reporting. This enables agents to work directly with enterprise data sources during extended tasks. Grok Bot: Grok Bot refers to SpaceXAI's persistent AI worker agents that maintain context and continue tasks after user sessions end. They emphasize dedicated computer environments for operating software. Recent expansions include shared Team Bots for organizational use with private conversations and common knowledge. SpaceXAI: SpaceXAI develops AI products including the Grok Bot line of persistent agents. It introduced Grok Bot agents in August and expanded with Team Bots in September, positioning them as long-running digital workers. The announcement highlights similarities and differences with Google's approach to agent identity and execution environments. Anthropic: Anthropic is an AI company that develops the Claude family of models. Its models are already supported by Google's new agent, and Anthropic recently announced native integration allowing Claude to work directly inside Google Docs, Sheets, and Slides. This partnership underscores the trend of cross-company model compatibility in enterprise agents. Snowflake: Snowflake provides a cloud data platform for data warehousing and analytics. Google's agent supports connections to Snowflake for data-related workflows. This allows persistent agents to incorporate data from multiple cloud sources without manual handoffs. Databricks: Databricks offers a platform for data engineering, analytics, and machine learning. It is among the tools integrated with the new Gemini agent for business operations. Such integrations help agents perform complex data and analytics work across systems. Salesforce: Salesforce provides customer relationship management and related enterprise software. Google's announcement lists Salesforce among the tools the new Gemini agent can connect to for workflow automation. This integration enables agents to access and act on data across business systems. ServiceNow: ServiceNow offers cloud-based platforms for IT service management and workflow automation. The Gemini agent is designed to integrate with ServiceNow as part of its broader tool ecosystem. Such connections allow persistent agents to handle tasks spanning multiple enterprise applications. Microsoft 365: Microsoft 365 is Microsoft's cloud-based productivity platform including Office apps, Teams, and related services. Google's new agent is designed to operate through Microsoft 365 interfaces in addition to Google's own tools. This multi-platform support reflects efforts to meet enterprises where they already work. Thomas Kurian: Thomas Kurian is the CEO of Google Cloud. He announced the new persistent Gemini agent and its capabilities for long-running tasks with dedicated identities and memory across systems. The announcement highlighted flexible model support and security controls managed through enterprise policies. Microsoft Teams: Microsoft Teams is Microsoft's collaboration platform for chat, meetings, and file sharing. The Gemini agent can operate through Microsoft Teams in addition to other interfaces. This broad compatibility supports deployment across mixed enterprise environments. Google Workspace: Google Workspace is Google's suite of productivity and collaboration tools including Gmail, Calendar, Drive, Docs, and Chat. The announcement positions the new Gemini agent to receive its own dedicated Workspace accounts, email addresses, calendars, and storage in one configuration for team use. This allows employees to interact with the agent through familiar interfaces much like assigning work to a colleague. Security: Agents receive cryptographically attested identities with permissions enforced through enterprise systems, audit logging, and an Agent Gateway for policy control. Competition: Several companies announced persistent or always-on AI agent features in the weeks surrounding Google's event, including updates from Microsoft, OpenAI, Anthropic, and SpaceXAI. Integration: The new Gemini agent is built to connect with a wide range of enterprise tools and data platforms while allowing enterprises to switch between different underlying AI models.