Atlassian is embedding artificial intelligence agents directly into its Jira project management software, marking a significant evolution in enterprise collaboration. The company has launched agents in open beta, allowing them to become proactive participants in workflows.
This move aims to address critical governance needs such as tracking work, auditability, privacy, and risk management. Jira's existing controls, including audit trails and role-based access, will extend to AI agents, ensuring that an agent's access mirrors that of a human user. Data integrity is further protected through private sandboxes, with agents unable to modify production code without human approval.
Atlassian's strategy is underpinned by its Teamwork Graph, which maps relationships between people, work, and knowledge across its platforms. This provides embedded agents with deep institutional knowledge, offering an advantage over standalone AI tools.
The expanded Rovo MCP Gallery now supports third-party agents from companies like GitHub, Box, and Figma. These integrated agents can access live data, perform actions like task creation, and combine external context with Atlassian's knowledge for improved reporting. Atlassian highlights that a substantial portion of these agent operations involve data writes, indicating active collaboration.
The company's broader objective is to prevent "agent sprawl" by managing, governing, and securing these AI tools effectively, ensuring a more unified and controlled AI integration within enterprise environments.