Atlassian has announced the open beta of AI agents in Jira, introducing AI capabilities into the environment where teams plan and manage their work. With this functionality, teams can assign tasks to Atlassian Rovo and third-party agents in Jira, interact with them through comments, and integrate them into their workflows. As a result, agent-driven work is no longer isolated and fragmented but visible, coordinated and integrated into key business processes.
Freedom of choice of AI agents
Atlassian has also announced new investments in Model Context Protocol (MCP), reinforcing its position as an open ecosystem that allows its customers to choose the agents and tools that best fit their organization’s needs.
“Work is changing rapidly: people now coordinate agents, tools and cross-functional teams. Without clear coordination, that complexity can turn into chaos,” says Tamar Yehoshua, Chief Product and AI Officer at Atlassian. «Our goal is to help teams transform that complexity into real productivity. “With these new capabilities, we bring agents into the tools and workflows our customers already know and trust, and provide them with an open, governed framework to integrate them as part of the team across the organization.”
AI Agents in Jira
Agents in Jira, now in open beta, allow teams to incorporate AI into their daily work through:
• Assigning tasks to Atlassian Rovo agents and MCP-enabled third-party agents.
• The mention of agents in the comments to collaborate iteratively and in context.
• Their integration into workflows so they can design, execute and update tasks, with people in control.
By operating within existing Jira structures, agents respect project settings, permissions, audit logs, and approval flows, allowing teams to confidently adopt AI.
An open ecosystem for AI agents, based on MCP
As organizations integrate new agent-based capabilities, Model Context Protocol (MCP) provides a common framework for AI agents to access tools, data, and workflows. Along these lines, Atlassian has announced new investments in this open ecosystem with the aim of integrating AI into the tools that teams already use.
With this integration, the company strengthens its position as a platform to coordinate work between human teams and AI agents
Adoption data among the company’s largest customers indicates that enterprises account for about 50% of total Rovo MCP server usage, while customers with paid editions account for 93% of usage. On this basis, Atlassian presents two new features:
• MCP skills available in Rovo: Rovo agents can connect to third-party applications compatible with MCP, such as Amplitude, Box, Canva, Figma or Intercom, to access real-time information, use their capabilities and execute actions.
• Rovo MCP Server general availability: This is a company-hosted MCP server that provides AI clients that support this standard with a single, secure point of connection to Jira and Confluence. These include Claude from Anthropic, Cursor, Gemini CLI from Google, Lovable and WRITER.
