Atlassian AI Features Included in Cloud vs Rovo Add-On: What You Get Free

The Confusion Between Atlassian Intelligence and Rovo

One of the most consistent sources of confusion in enterprise Atlassian licensing is the distinction between Atlassian Intelligence — the native AI capabilities embedded in Atlassian Cloud products — and Rovo, the broader AI platform. Many organisations renewing Cloud subscriptions or evaluating migration from Data Center to Cloud are uncertain about what AI functionality they are receiving as part of their standard subscription and what requires additional cost through Rovo's credit system. This confusion has commercial consequences, because it affects both the budget planning for AI adoption and the negotiation strategy for Atlassian contract renewals.

Understanding this distinction precisely is also essential for the Rovo AI licensing negotiation, because the credit consumption model only applies to capabilities beyond the native Intelligence features. Getting this boundary clear before entering commercial discussions prevents both overspending on Rovo access for needs that are already met by native features and underspending on Rovo capabilities that deliver material value beyond what native Intelligence provides.

What Atlassian Intelligence Includes at No Extra Cost

Atlassian Intelligence is the set of AI capabilities embedded directly into Atlassian Cloud products and available to all Cloud subscribers without credit consumption or Rovo-specific licensing. These features are woven into the native product interface rather than accessible through a separate AI platform. They represent Atlassian's baseline investment in making its products more intelligent — and they are included in Standard, Premium, and Enterprise Cloud subscriptions without additional cost.

In Jira, Atlassian Intelligence native features include AI-powered issue summarisation that condenses lengthy issue descriptions and comment threads into concise summaries, automated sprint retrospective insights that identify patterns across completed sprint work, smart assignment recommendations based on historical team patterns, and natural language to JQL conversion that allows users to construct Jira filters by describing what they want in plain language rather than learning Jira Query Language syntax. These features require no credit consumption and are available to all users on the subscription.

In Confluence, native Atlassian Intelligence provides AI writing assistance that helps users draft, refine, and expand page content within the editor interface, page summarisation that condenses long documents for quick review, AI-powered suggestions for action items and meeting notes within structured templates, and content transformation tools that convert bullet points to prose or restructure content on request. Again, these features are part of the standard Cloud subscription and carry no per-use cost.

In Jira Service Management, native features include smart quick replies that suggest response drafts for service agents handling customer requests based on previous similar responses, automated categorisation of incoming requests, and AI-assisted knowledge base suggestions that surface relevant Confluence articles during ticket resolution. These capabilities improve ITSM team productivity without requiring Rovo access or credit consumption.

The Atlassian Intelligence native features represent meaningful productivity improvements for the average Atlassian Cloud user. For organisations primarily concerned with reducing manual work in standard Jira and Confluence workflows, the native features may be sufficient — and ensuring this is understood before purchasing additional Rovo capacity is an important step in avoiding unnecessary AI licensing spend.

What Rovo Adds: Beyond Native Intelligence

Rovo extends the AI capability surface in ways that native Atlassian Intelligence does not — and the distinction is fundamentally about scope, depth, and automation capability rather than quality of the underlying AI. Rovo operates across the full Atlassian ecosystem and connected external platforms rather than within individual product interfaces, and it enables agentic automation that executes multi-step workflows rather than just assisting with individual tasks within a product.

Rovo Search is the most frequently encountered Rovo capability and, critically, it is unlimited and does not consume credits. Rovo Search queries across Atlassian tools and more than 100 connected external platforms — Google Drive, Microsoft SharePoint, Slack, GitHub, Salesforce, and others — in a unified interface. The results are contextually ranked using the Atlassian Teamwork Graph, returning genuinely relevant results rather than keyword matches across siloed systems. This capability is meaningfully different from the search experiences native to individual Atlassian products and represents immediate value for organisations with knowledge distributed across multiple systems. Since it carries no credit cost, it should be considered effectively free within Cloud subscriptions.

Rovo Chat is where credit consumption begins. Each Rovo Chat interaction costs 10 credits from the monthly pool. Rovo Chat differs from native product AI in that it operates across the entire organisational data estate rather than within the scope of a single product. A Rovo Chat query can synthesise information from a Jira project, the associated Confluence space, the JSM service history for a related customer, and the Slack channel where the work was discussed — all in a single response with organisational context. This cross-system synthesis is the capability that native Atlassian Intelligence does not provide.

Rovo Agents are customisable automations that execute multi-step tasks within and across Atlassian tools without requiring human intervention at each step. Standard Rovo Agent interactions consume 10 credits. Unlike native Jira Automation, which executes rule-based automations triggered by specific events, Rovo Agents operate with AI-level reasoning about what steps are required to accomplish a goal — they can handle ambiguity, make conditional decisions, and execute workflows that would require human judgment to navigate in a traditional automation framework. Deep Research, at 100 credits per request, is the most resource-intensive Rovo operation — it synthesises information across multiple sources and constructs comprehensive responses to complex research questions, consuming significantly more compute than standard Rovo Chat or Agent interactions.

The Credit Consumption Model in Practice

Understanding how the credit pool translates to practical usage is essential for budget modelling and for negotiating appropriate credit allocations in Atlassian Cloud contracts. The monthly credit pool is allocated at the plan level — Premium Teamwork Collection provides approximately 70 credits per user per month, pooled across the organisation; Enterprise provides approximately 150 credits per user per month.

In practical terms, a Premium user with 70 monthly credits can conduct approximately seven standard Rovo Chat interactions, seven standard Rovo Agent executions, or some combination of both before exhausting their share of the pool. A single Deep Research request consumes 100 credits — more than the entire monthly share of an individual Premium user. This makes Deep Research a capability that must be used selectively and should be assessed carefully in credit modelling before users have unlimited access to it without awareness of the credit cost.

For organisations modelling whether the standard credit allocation is sufficient, the key variable is the proportion of users who will be active Rovo Chat and Agent users versus those who will primarily use Rovo Search and native Atlassian Intelligence. In most enterprise deployments, a significant proportion of the user base will be passive Rovo users — they benefit from native Intelligence features and Rovo Search, but do not regularly initiate Chat or Agent interactions. The credit pool accumulates from these passive users and is available for active users, making the pooled model more generous than a per-user allocation would suggest.

Practical Guidance: When Native AI Is Enough

For organisations evaluating whether to invest in Rovo beyond the free Rovo Search tier, the decision hinges on several factors. The first is knowledge distribution: if the organisation's work knowledge lives primarily within Atlassian tools, native Intelligence features are more likely to meet the synthesis and assistance needs. If knowledge is distributed across Atlassian, Google Workspace, Slack, GitHub, and other platforms, Rovo's cross-system reach delivers value that native Intelligence cannot.

The second factor is automation ambition. If the primary AI use case is assisting individual users with writing, summarisation, and search, native features combined with free Rovo Search may be sufficient. If the ambition extends to agentic automation of multi-step processes — ticket routing, knowledge base maintenance, cross-system workflow orchestration — Rovo Agents are required and the credit model becomes relevant.

The third factor is user sophistication. Native Atlassian Intelligence features are designed for passive use — they surface in context without requiring users to change behaviour. Rovo Chat and Agents require active user engagement and an understanding of how to prompt effectively for the best results. Organisations without an AI adoption programme that trains users on effective prompting patterns will see lower utilisation of Rovo Chat and Agent capabilities and correspondingly lower return on credit investment.

Connecting This to the Atlassian Contract

The native AI vs Rovo distinction has direct implications for Atlassian contract negotiation. When renewing or renegotiating Cloud contracts, organisations should explicitly request documentation of which AI capabilities are included in the base subscription at no additional cost versus which require credit consumption. This documentation should be attached to the contract as a schedule, rather than existing only as reference to current product documentation that Atlassian can update at any time.

The Cloud contract negotiation process should address credit pool sizes explicitly — the contract should specify the monthly credit pool numerically rather than by reference to the plan-level policy. It should also specify the conditions under which Atlassian can change the credit consumption rate for specific Rovo capabilities. Atlassian's policy commitment to 90 days' notice before enforcing credit changes is not a contractual commitment unless it is embedded in the signed agreement.

The pricing changes Atlassian introduced in 2025 and 2026 have already demonstrated that the company is willing to restructure the value-for-money equation of its Cloud subscriptions. Protecting the credit model terms contractually is a prudent measure for organisations making multi-year AI adoption investments on the Atlassian platform.

Our Atlassian contract advisory team reviews AI licensing terms, models credit consumption against usage patterns, and negotiates the contract protections that enterprise AI investments require. Contact us to ensure your Atlassian Cloud contract reflects the AI capabilities you are using and protects against future commercial changes to the credit model.