Atlassian introduces AI agents in Jira for enhanced automation and workflow integration

Atlassian has introduced AI agents in Jira, bringing a new level of automation now available in open beta. Teams can assign work directly to agents or mention them in comments to prompt iteration or get specific follow-up actions. Agents display as assignees with standard Jira fields, making it easy for teams to recognize tasks handled by agents while tracking each item’s status within sprints, releases, or incidents.

Following this introduction, administrators are given fine-grained control to determine which agents are accessible to which users, in what contexts they can be used, and the specific criteria for marking tasks as complete. Individual exchanges with agents, including requests, agent responses, and final team decisions, are preserved in the relevant work item for complete transparency.

Agents can be embedded into Jira workflows, allowing them to perform certain jobs automatically within existing processes. In addition, agents adhere to Jira’s established project settings, permissions, and auditing features, ensuring that organizational standards are maintained. Because Atlassian’s Model Context Protocol (MCP) creates an open and standardized interface, users can connect custom models and agents to Jira, integrate with proprietary tools and data, and configure agent actions directly within their workflows.