If your team uses Jira, Confluence, Trello, or any other Atlassian product, you should know this: as of this month, Atlassian quietly enabled default data collection across its platform to train AI.
No big announcement. No prominent notification. A policy change, an email to admins, and a promise that opt-out controls are coming in May.
This is not a hypothetical. It is happening right now.
What Changed
Atlassian updated its data handling policies to allow the company to collect content from your Atlassian products, including tickets, docs, and project data, to train and improve its AI features (collectively called Atlassian Intelligence and Rovo).
The change was surfaced this morning on Hacker News, where it quickly gained nearly 200 upvotes and dozens of comments from engineers and business owners who had not seen any prominent warning before the change took effect.
According to a comment from a user who received an email from Atlassian: "They said the opt-out features will be rolled out to the Admin portal in May."
So for now, if you are an Atlassian customer, you are in by default.
Why This Matters for Small Businesses
Big companies have dedicated legal and security teams who catch policy changes like this. Small businesses usually do not.
If your team uses Jira to track client work, Confluence to store internal docs, or Trello to manage projects, the content inside those tools may now be used to train AI models. That could include:
- Client names and project details
- Internal pricing or strategy documents
- Support tickets with customer information
- Proprietary process documentation
Whether this violates your client contracts, NDAs, or data agreements depends entirely on what is in your Atlassian workspace. But the decision was made for you, not by you.
What You Should Do Right Now
1. Check your Atlassian admin settings today. Log in as an admin and look under Privacy or Trust settings. Some controls may already be visible, even if the full opt-out portal is not live until May.
2. Review your data. Take 10 minutes to think about what sensitive information lives in your Atlassian tools. Client data, financial details, anything under NDA. If it is there, flag it internally.
3. Get on the opt-out list. When Atlassian rolls out the opt-out controls in May, make opting out the first thing you do. Set a calendar reminder. Do not wait for it to appear on its own.
4. Communicate with your team. If team members are adding sensitive client content to Confluence or Jira, they should know this policy changed. A quick Slack note or email is enough.
5. Talk to legal if it matters. If your business handles sensitive client data under contract, loop in your legal contact or review your client agreements. Better to check now than explain later.
The Bigger Pattern
Atlassian is not alone. Over the last two years, nearly every major SaaS platform, including Zoom, Slack, Google Workspace, and Microsoft 365, has updated its policies to allow some form of AI training using customer data. Most have faced similar backlash and most eventually added opt-out controls.
The pattern is consistent: companies opt you in first, then offer controls after the reaction.
The lesson is also consistent: small businesses cannot assume the defaults are in their favor. With AI now embedded in nearly every tool you use, data collection is becoming a standing agenda item, not a one-time thing to check.
Atlassian's products are used by millions of small teams worldwide. If your company is one of them, this week is a good week to take 20 minutes and find out exactly what they have on you and what you can turn off.
Source: Hacker News discussion (April 20, 2026), Atlassian community emails reported by users. Opt-out controls expected in Atlassian Admin portal in May 2026.