Atlassian is going all in on AI training data. Starting August 17, 2026, the company will use customer metadata and in-app data to improve AI features across its platform, including Rovo and Rovo Dev. Settings rollout begins April 16, 2026, giving organizations time to prepare.
So what's actually being collected? Metadata covers things like issue types, project structures, and usage patterns. In-app data is broader: comments, descriptions, content you create in Jira, Confluence, and other Atlassian tools. The company says all contributed data gets de-identified and aggregated before use. In practice, that means stripping names, emails, and other identifiers, then grouping data so nothing traces back to a specific person or org.
The defaults depend on your plan. In-app data collection turns on by default for Free and Standard plans, though admins can opt out. Metadata collection defaults to ON for everyone. If you're on a cheaper plan, you're sharing metadata whether you like it or not unless you upgrade to Enterprise.
Microsoft, Salesforce, and Google are all doing the same thing. Atlassian frames this as responsible AI with privacy built in, and they're hosting a webinar on April 28, 2026 to walk customers through the changes.
For teams on Free or Standard plans, the math is simple. Check your org settings before August 17 or accept the defaults.