OpenAI models are now behind Atlassian's Rovo agents

Atlassian and OpenAI expanded their partnership, bringing frontier models from the GPT-6 family across Atlassian's platform. Under the new agreement, OpenAI models will power agents across Atlassian's platform and Rovo, and the companies framed the deal as a way to turn enterprise knowledge into action.

The collaboration dates back to 2023. This expansion is partly driven by Atlassian's own internal adoption: more than 3,000 Atlassian developers use Codex across terminals, IDEs, and code review workflows, per the announcement, and OpenAI continues to use Jira to manage its own critical internal workflows.

What the deal covers

The core mechanism is the Teamwork Graph, Atlassian's enterprise context layer that connects people, projects, documents, and decisions. Rovo combines OpenAI models with that graph, so the agents get context about how a company actually works instead of general knowledge alone.

Atlassian also gets expanded access to the latest OpenAI frontier models, including GPT-6 Astra and the GPT-5.6 series, as the models advance. The access runs through OpenAI's API, so Rovo picks up new reasoning capabilities as OpenAI ships them, according to the announcement.

What it looks like in practice

The example the companies give: a product manager asks Rovo whether a launch is on track. Rovo pulls together Jira tickets, Confluence documents, and the related discussions, identifies engineering blockers, flags missed milestones, and surfaces decisions that need attention. The model turns all of that into a readiness assessment with recommended next steps.

Beyond Rovo, Atlassian's ChatGPT and Codex plugins connect those tools to existing workflows. The plugin extension puts Jira work items, Confluence content, and people directly into prompts, and the pinned Atlassian Home surfaces assigned work, recent Looms, projects, and Bitbucket pull requests. Everything stays subject to the user's permissions.

Where it heads next

The companies say they are exploring deeper Jira integrations: assigning work to AI agents, tracking progress, capturing decisions, and reviewing results. Paired with DX, Atlassian's platform for measuring developer productivity, that would let engineering leaders measure AI's impact on development speed and cycle time while keeping humans in control.

For a team already running Jira and Confluence, the practical question is where your knowledge lives. The deal puts frontier models where the project context already sits, so the value shows up if your documentation and decisions are in those tools. If most of your team's context is scattered across chat and email instead, this changes little for you, at least until the agent integrations start pulling from there.