Atlassian expands Jira to orchestrate developers and AI agents

Atlassian announced a suite of updates to Jira designed to position it as an orchestration hub for mixed teams of human developers and AI agents. The updates include three main components: Jira Planner, which transforms incomplete project ideas into technical specifications; Jira Coding Agent, which can be triggered directly from Jira to handle bounded work in the cloud; and integrations with third-party agents such as Claude, Cursor, and GitHub Copilot.

Central to the announcement is the Agentic Engineering Template, aimed at advanced adopters. According to Ming Wu, Head of Engineering for DevAI, the template works by setting up a Jira board where columns represent workflow states. When an issue moves from one column to another, agents are automatically assigned, creating a level of detailed and nuanced automation. This design allows developers to delegate routine tasks without interrupting their own work, while keeping all agent activity grounded in Jira's project history and knowledge base.

The motivation for these changes comes from a recognition that broad coding-agent adoption has not automatically delivered value. Research from Queen's University Kingston analyzing 61,000 repositories and 47,000 developers found that AI agent submissions are accepted less frequently than human-authored code and tend to be structurally simpler. Wu described this gap as a problem of coordination and context. Even as frontier AI models get smarter and faster, agents struggle without clear requirements, project context, proper handoffs, environment setup, and governance—what Atlassian calls the "work that surrounds work." The company believes the bottleneck is not agent capability but the lack of a cohesive solution to organize, prioritize, and track work across a mixed human-and-agent team.

Atlassian's strategy is to avoid positioning Jira as merely another tool and instead frame it as a solution spanning the entire software development lifecycle. Wu stated: "We need a solution rather than a tool... a solution across the entire software development lifecycle journey to actually address different pain points." By integrating agents from multiple vendors and automating assignment workflows, Atlassian aims to help companies realize return on investment from AI tool spending—a shift from simply boosting agent usage metrics to delivering measurable business value.

Atlassian's expansion of Jira reflects a broader recognition that the AI coding boom has hit a maturity plateau. While individual coding agents (such as GitHub Copilot and Claude) have seen rapid uptake, companies deploying them are discovering that raw code generation speed does not translate to project value. Research cited in the announcement—a study by Queen's University Kingston reviewing 61,000 repositories and 47,000 developers—found that AI agent submissions are accepted less frequently than human code and tend to be structurally simpler. This gap appears to stem not from agent capability alone, but from friction in the orchestration layer: agents lack clear requirements, project context, and integration with human workflows.

Atlassian's bet is that Jira, already the dominant project-management platform for software teams, can evolve into the control plane where developers and agents coordinate. By automating task assignment, providing context from project history, and embedding support for third-party agents (not just Atlassian's own Jira Coding Agent), the company aims to unlock the value trapped in isolated agent tools. Head of Engineering Ming Wu emphasized this strategy explicitly: "You want to bring the true value rather than boosting usage itself"—a clear signal that Atlassian is targeting return on investment, not just adoption metrics.

No comments yet. Be the first to share your thoughts!