Atlassian Expands Jira With Autonomous AI Agents Built for Continuous Development

Atlassian plc is rolling out fresh Jira functionality aimed at enabling engineering organizations to operate AI agents at scale while maintaining control over their actions throughout extended periods. The shift toward always-on agentic artificial intelligence represents the next frontier in how software gets built, with teams increasingly deploying agents not just to handle more tasks, but to work autonomously for longer stretches and across more stages of development.

As organizations expand their use of AI agents, they face new challenges around trust, grounding, shared context, communication, institutional memory and validation. Atlassian's response includes a suite of updates designed to manage agent behavior, establish operational standards, monitor active agents, and confirm that usage patterns remain sound.

One core issue Atlassian identified is that agents frequently stumble when they lack sufficient understanding of project architecture, prior decisions and established standards. The company is addressing this through Code Context, a capability built on its Teamwork Graph technology that furnishes coding agents with secure access to intelligence spanning intricate, multi-repository codebases.

Complementing this foundation are Agent Space Settings and Agent Context Controls, which define the boundaries of where agents operate, what information they can access and which actions they can take. This approach mirrors how teams manage human employee permissions, giving both team members and management the ability to enforce security policies consistently.

The vision for modern AI agents differs fundamentally from chatbots that require constant user direction. Instead, these systems can execute independently for extended durations—potentially hours or days—activating themselves when tasks arise and reporting back on progress.

Atlassian has introduced Agent loops within Jira, a mechanism that examines work backlogs, converts items into code merge requests, applies a standards-checking system to ensure code conformance, and deploys a dedicated agent to evaluate merge requests against those standards and surface any violations.

Automatic but not invisible

These autonomous operations remain fully transparent and subject to validation. Since no universally accepted playbook exists yet and industry best practices continue to evolve, Atlassian provides accountability and visibility infrastructure on the backend to document activity.

Each agent execution produces an audit log containing diagnostic information, enabling the system to display and quantify AI's contribution across metrics like throughput, quality, adoption and cost. Teams gain visibility into spending as code moves toward production. The offering also includes an AI agent usage dashboard that reveals who is leveraging the tools, in what manner, and what results emerge at the team level.

Atlassian's goal is to help teams transition from relying solely on foreground coding tools to embracing always-on agents that operate silently in the background, handling repetitive engineering work. Today, many organizations are exploring these capabilities in an unstructured way, without adequate frameworks to govern, integrate or measure their impact.