Does Qwen train on your data?
Qwen's policy says it trains its AI models only on de-identified content and feedback, never on raw inputs, and that people do not read what you submit. The way out is a request to privacy@qwen.com.
"We may use de-identified user content and feedback data to train and improve the accuracy, performance, and quality of our AI models and services. In such cases, we use only de-identified data that cannot reasonably be linked to a particular individual."
- 1 Oct 2026
- 5 Oct 2026
- All accounts (qwen.ai and chat.qwen.ai)
- global
- Fetched, browser-rendered
- 68B646D7…CE3E
How we read it: Re-read on 22 September 2026 after the policy was rewritten on 21 September, ten days after this site's letter of 11 September, which drew no reply. The row this entry quoted is gone. The policy now says training uses de-identified content and feedback only, that "we will not use your raw input data to directly train AI models", and that "we will not manually review the input content you submit" except for security or legal reasons, so human review moves from unknown to no. The controller is now NTH POWER GLOBAL TECH SINGAPORE PTE. LTD., with privacy@qwen.com for rights requests. The way out is still a request under the right to object; no setting is named, so the grade stays a D.
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.
- Contact support
- Not stated
- 49 / 100
- Yes, stated
- No
How to opt out of Qwen AI training
Plan by plan, from what Qwen'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 accounts (qwen.ai and chat.qwen.ai)
Yes, but not self-serve: Qwen makes you contact support.
What this policy has done since we started watching
2 changes detected since 6 Sept 2026. Each one is a diff between two archived copies, and each links to the passage that moved.
Everything we read
Vendor response
We wrote to Qwen (Alibaba) on 11 Sept 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