How Rovo Agents Are Transforming JSM in 2026

Jira Service Management has always been a powerful platform for IT teams, support teams, and operations teams managing complex service environments. But managing alerts, handling on-call schedules, triaging incoming requests, and keeping Jira boards current has still required a significant amount of manual coordination. Someone has to watch the queue. Someone has to acknowledge the alert. Someone has to update the ticket after the meeting.

Rovo Agents are changing that. Across a series of updates in 2026, Atlassian has deepened the integration between Rovo and Jira Service Management in ways that go significantly beyond a virtual chatbot answering employee questions. Rovo Agents can now be embedded directly into JSM portals and help centers, connected to on-call schedules and incident queues through the Rovo MCP Server, and configured to create structured JSM requests from natural language conversations using real request types and forms. For IT and operations teams, this represents a meaningful shift in how service management actually works day to day.

Rovo Agents in the JSM Portal and Help Center

One of the most visible changes in 2026 is how accessible Rovo Agents have become for the people actually using JSM to submit requests and get help. Rovo Agents can now be added directly to JSM portals with a single click. Once active, the agent sits inside the portal as a chat widget, drawing on the organization's existing knowledge base to answer questions and resolve issues before a ticket is ever submitted.

When the agent cannot resolve an issue on its own, it escalates to a human agent by creating a JSM ticket automatically, passing along the full context of the conversation so the support team does not have to ask the employee to repeat themselves. For organizations that have been frustrated by the limitations of traditional virtual service agents, the difference in capability is significant. Rovo Agents use generative AI to handle complex, open-ended conversations rather than rigid decision trees, which means they can handle a much wider range of employee requests without breaking down into unhelpful responses.

Rovo Agents are now accessible to all Atlassian account holders, including internal users who do not hold a paid product license. This closes a major adoption gap that previously limited self-service AI in JSM environments to licensed users only.

Smart JSM Request Creation From Natural Language

A long-standing friction point in service management is the gap between how employees describe a problem and what information IT actually needs to resolve it. Traditional request forms help structure that information but they require the employee to navigate form fields, understand field requirements, and fill in everything correctly before submitting. When they do not, the ticket arrives incomplete and the back-and-forth begins.

Rovo Agents now handle this intelligently. When a user initiates a conversation with a Rovo Agent and describes their issue in plain language, the agent uses that conversation to intelligently populate the relevant JSM request type and form fields. It works with any request type regardless of complexity, handles cascading selects, custom fields, date fields, and other complex field types automatically, and allows users to adjust the request conversationally before it is submitted.

A user can say something as simple as "I need a new laptop" and the agent will identify the correct request type, populate all required fields based on the conversation, and confirm the details with the user before submitting. For IT teams managing high-volume request queues, this means fewer incomplete tickets, fewer clarification requests, and faster resolution from the moment a request arrives.

Rovo MCP Server Now Supports JSM Tools

The most significant capability expansion for operations and DevOps teams in 2026 is the addition of JSM-specific tools to the Atlassian Rovo MCP Server. The MCP Server allows AI assistants and development tools to connect directly to Atlassian products, enabling natural language interaction with Jira, Confluence, Compass, and now Jira Service Management without switching context.

With JSM tools added to the MCP Server, AI agents and developers can now query JSM alerts by ID or search criteria, view on-call schedules and identify who is currently on call or on call next, acknowledge, close, or escalate alerts, get team information including escalation policies and roles, and interact with incident queues directly from their IDE or AI assistant.

In practice this means an engineer responding to an incident no longer needs to switch between their development environment and Jira Service Management to check who is on call, acknowledge an alert, or update an incident status. They can ask their AI assistant directly and get an answer or trigger an action without leaving the tool where they are already working. For teams where minutes matter during incidents, eliminating that context-switching has a direct impact on response times.

One important note is that JSM tools in the MCP Server are currently available using API token authentication only and are not yet supported via OAuth. Organizations that have the Atlassian MCP installed with OAuth will need to also install it using the API token method to access the JSM tools.

Two-Way Slack and JSM Comment Syncing

For teams that run incident response and service management discussions in Slack, one of the most practically valuable updates is the introduction of two-way comment syncing between Slack threads and JSM tickets through Rovo Agents.

When a JSM ticket is raised from a Slack thread using Rovo, all messages from the original thread are automatically added as comments on the ticket. From that point forward, new public comments added to the JSM ticket appear in the linked Slack thread, and replies in Slack sync back to JSM. This works in both directions and supports edits, deletions, and image attachments. Internal comments in JSM remain private and are not synced to Slack.

For teams that have historically struggled with the gap between where incidents are discussed and where they are tracked, this update means the two environments stay in sync automatically without anyone having to manually copy information from one place to the other.

Proforma Forms for Richer Request Creation

Beyond smart natural language request creation, Rovo Agents now support raising JSM requests that include Proforma forms rather than just basic request fields. This allows organizations to collect richer, more structured information from users at the point of request creation directly inside the agent conversation.

For teams that use Proforma forms to capture detailed information for specific request types, such as onboarding checklists, procurement requests, or access requests, this update means those structured forms are now accessible through a Rovo Agent conversation rather than requiring the user to navigate to the portal and fill them in manually.

Assigning Work Items Directly to Rovo Agents

One of the most fundamental changes to how JSM workflows operate is the ability to assign work items directly to Rovo Agents, exactly the same way you would assign a work item to a human team member. This allows organizations to route specific categories of work to agents automatically as part of their JSM workflow, rather than relying on automation rules to trigger agent actions after the fact.

An agent assigned to a ticket can triage the request, gather missing information, update fields, route the ticket to the right team, and in many cases resolve it entirely without human intervention. For IT teams handling large volumes of routine requests, this represents a meaningful reduction in the manual triage work that currently consumes a disproportionate share of service desk capacity.

Usage Analytics and Performance Reporting for JSM Agents

Understanding whether your Rovo Agents are actually working is now significantly easier with the addition of resolution and deflection reporting inside Rovo Agent Studio. Admins can access metrics on how effectively agents are resolving issues and deflecting tickets, see trends and patterns in agent conversations, and understand how many users are engaging with each agent over time.

For JSM leaders who need to demonstrate the value of AI investment to leadership or justify expanding agent deployments, these metrics provide the concrete evidence that conversations about ROI require. Resolution rates and ticket deflection numbers are exactly the data points that make the business case for continued AI adoption in service management environments.

What This Means for IT and Operations Teams

Taken together, the Rovo Agent updates for JSM in 2026 represent a meaningful change in what AI-powered service management can actually do. The early phase of Rovo in JSM was primarily about answering employee questions and deflecting simple tickets. These updates move significantly beyond that baseline.

Agents are now embedded where work happens.

Rather than sitting as a separate tool employees have to seek out, Rovo Agents live directly inside JSM portals, help centers, and Slack channels. Help is available in the same place the employee already is, which means higher adoption and fewer tickets that go unresolved simply because someone could not find the right channel.

Complex requests no longer require perfect form-filling.

Employees describe what they need in plain language and the agent handles the rest, populating the correct request type, filling in the required fields, and confirming the details before submission. IT teams receive complete, structured tickets instead of half-filled requests that require follow-up before work can even begin.

Incident response is faster and less fragmented.

Engineers managing live incidents can now query on-call schedules, acknowledge alerts, and update incident status through their AI assistant without switching tools. The back-and-forth between environments that previously added minutes to every incident response is gone.

Slack and JSM stay in sync without anyone maintaining them.

Two-way comment syncing means decisions made in Slack appear in JSM automatically and updates made in JSM flow back to Slack. The gap between where incidents are discussed and where they are tracked closes itself.

For IT leaders evaluating whether to expand their Rovo deployment into JSM, the capability gap between what Rovo could do a year ago and what it can do now is substantial.

The Role of Atlas Bench

Deploying Rovo Agents effectively in a JSM environment requires more than activating features. It requires configuring agents around your specific request types and workflows, connecting the MCP Server correctly for your on-call and incident management setup, setting up two-way Slack syncing in a way that matches how your teams actually communicate, and ensuring the right governance frameworks are in place so agents operate reliably at scale.

As an Atlassian Platinum Solution Partner and World Class Software Development Finalist, 2024-2025, Atlas Bench helps IT and operations teams implement Rovo Agents in ways that deliver real, measurable improvements to service delivery. We configure the platform around how your teams actually work and stay involved beyond the initial deployment to ensure agents perform as expected as your environment evolves.