Does Atlassian train on your data?
Atlassian uses support interactions to train AI models — Jira/Confluence content isn't addressed here. No opt-out is stated outside the EU/UK, where objecting may mean no longer using the Services.
No opt-out found in this policy
Graded on the plan most people are on. The other plans are in the matrix below.
"We use your information to resolve technical issues you encounter, to respond to your requests for assistance, to analyze crash information, and to repair and improve the Services, including for development, training, or fine-tuning of machine learning and artificial intelligence models."
- 1 Sept 2026
- 5 Oct 2026
- All plans — customer support interactions
- global
- Fetched from the company
- 81A74B8E…4DD3
How we read it: Scoped narrowly: this covers information you provide through Atlassian's support channels (bug reports, crash info, requests for help) — not the content inside your Jira issues, Confluence pages, or other product data, which this privacy policy does not clearly resolve either way (see the Product content tier below, graded UNCLEAR — not confirmed excluded). Outside the EU/UK, no opt-out is stated for this specific use anywhere in the document; the only global channel is "ask us to stop accessing, storing, using and otherwise processing your information where you believe we don't have the appropriate rights to do so" — conditioned on a rights dispute, not a targeted opt-out.
opt out, opt-out, object to, objection, stop processing, withdraw, settings, preferences
What the document does offer
- the right to object to processing done on legitimate interest — the basis named for AI/ML training — but the section it sits in opens "If you are an individual in the European Economic Area (EEA) or United Kingdom (UK)", and the policy warns objecting "may mean no longer using the Services"
- withdraw consent, where consent was the basis for a specific purpose
- opt out of targeted advertising, via "Manage Preferences" or a Global Privacy Control signal
- control or opt out of cookies and tracking technologies, per the Cookies & Tracking Notice
- opt out of promotional communications, in the message itself or in account settings
Why none of it applies: The objection right is the only control that reaches model training, and the document gates it by geography rather than by plan. Everything else covers advertising, cookies, promotional email, or a consent that was never the basis for this use. So the absence claimed in the headline is the absence outside the EEA/UK, which is what the headline says.
This clause is real, but it doesn't resolve the training question - see the finding below.
"The Services include the Atlassian products you use, where we collect and store content that you post, send, receive and share. We process this content in the role of a processor or service provider on behalf of our customers."
- 1 Sept 2026
- 5 Oct 2026
- Product content (Jira, Confluence, Trello, etc.)
- global
- Fetched from the company
- 81A74B8E…4DD3
Why this is still UNCLEAR: The privacy policy states that content you post, send, receive, and share through Atlassian's products is processed "in the role of a processor or service provider on behalf of our customers" — governed by the Data Processing Addendum Atlassian signs with the customer (typically your employer), not by this privacy policy directly, and not gated by which plan you're on. That DPA isn't a source this monitor watches. The contract that sat beside it, the Cloud Terms of Service, was replaced on 17 August 2026 by the Atlassian Customer Agreement at the same address; until 1 September this entry was still quoting the old agreement's licence over "Your Data", a clause that no longer exists, and the fault for that was ours. The new agreement routes the question twice rather than answering it: "Atlassian may process Customer Data as specified in Section 6 (Purposes of the Processing) of Schedule 1 of the DPA", and "Atlassian provides AI Offerings under the AI Terms". So this site now reads the AI Terms too. They establish that "Input and Output are Customer Data under the Agreement", and they carry one real commitment: "Atlassian must not permit its subcontractors to use Input or Output to train or improve their models". That is a bar on other people's models. On Atlassian's own, they say only that "The Documentation provides more information about Atlassian's use of Input and Output and the privacy preserving measures Atlassian applies" — a fourth document, unwatched. Separately, the privacy policy's "develop and improve our Services" purpose (the same purpose the legal-basis table ties to ML/AI model training) mentions using "content that has been de-identified and aggregated in accordance with applicable data contribution product functionalities" — a real, if vague, pipeline from product content toward Service development that we can't confirm reaches AI training specifically, is opt-in, or is opt-out. Three documents read, each pointing onward; genuinely unclear on the evidence archived, not a lack of looking.
What each plan can refuse
The same questions, asked of every plan we could identify. Every cell comes from a quoted clause or a documented search - see the evidence above.
- None offered
- Not stated
- 87 / 100
- Not stated
- Unknown
- Not established
- Not stated
- —
- —
- Not established
How to opt out of Atlassian AI training
Plan by plan, from what Atlassian's own policy says about refusing. Where it publishes the steps, they're listed here with the limits they carry. Where there's nothing to switch off, or nothing graded, this says so instead of inventing a menu path.
All plans — customer support interactions
No opt-out is stated in the policy on file for Atlassian.
Product content (Jira, Confluence, Trello, etc.)
Not graded, so no route is pointed at. The privacy policy states that content you post, send, receive, and share through Atlassian's products is processed "in the role of a processor or service provider on behalf of our customers" — governed by the Data Processing Addendum Atlassian signs with the customer (typically your employer), not by this privacy policy directly, and not gated by which plan you're on. That DPA isn't a source this monitor watches. The contract that sat beside it, the Cloud Terms of Service, was replaced on 17 August 2026 by the Atlassian Customer Agreement at the same address; until 1 September this entry was still quoting the old agreement's licence over "Your Data", a clause that no longer exists, and the fault for that was ours. The new agreement routes the question twice rather than answering it: "Atlassian may process Customer Data as specified in Section 6 (Purposes of the Processing) of Schedule 1 of the DPA", and "Atlassian provides AI Offerings under the AI Terms". So this site now reads the AI Terms too. They establish that "Input and Output are Customer Data under the Agreement", and they carry one real commitment: "Atlassian must not permit its subcontractors to use Input or Output to train or improve their models". That is a bar on other people's models. On Atlassian's own, they say only that "The Documentation provides more information about Atlassian's use of Input and Output and the privacy preserving measures Atlassian applies" — a fourth document, unwatched. Separately, the privacy policy's "develop and improve our Services" purpose (the same purpose the legal-basis table ties to ML/AI model training) mentions using "content that has been de-identified and aggregated in accordance with applicable data contribution product functionalities" — a real, if vague, pipeline from product content toward Service development that we can't confirm reaches AI training specifically, is opt-in, or is opt-out. Three documents read, each pointing onward; genuinely unclear on the evidence archived, not a lack of looking.
What this policy has done since we started watching
11 changes detected since 1 Jan 2023. Each one is a diff between two archived copies, and each links to the passage that moved.
- Touched opt-outs, retention, what's collected, sub-processors, plan tiers and training.
- MajorTouched jurisdiction, opt-outs, retention, what's collected, sub-processors and training.
- Touched jurisdiction, opt-outs, retention, what's collected, sub-processors, plan tiers and training.
- Touched jurisdiction, opt-outs, what's collected and sub-processors.
- Touched opt-outs, retention, sub-processors and training.
- MajorTouched jurisdiction, opt-outs, retention, sub-processors and plan tiers.
- Touched opt-outs and sub-processors.
- MajorTouched jurisdiction, opt-outs, retention, what's collected, sub-processors, plan tiers and training.
Vendor response
We wrote to Atlassian (Jira / Confluence / Trello) on 20 Aug 2026, quoting the clauses above and inviting a correction. Nothing has come back. That's published because a company declining to comment on its own policy is something a reader should know, and because the invitation should be on the record whether or not it's taken up. Any reply goes here, word for word.
Report a wrong clause