Does Character.AI train on your data?
Character.AI trains on the conversations you submit. You can opt out — but the right is stated in a regional supplement, not the privacy policy, and points at a FAQ.
A self-serve opt-out exists
"With specific regard to the processing of your personal data for the purpose of training and improving our AI models, you have the right to opt out at any time."
- 17 Aug 2026
- 5 Oct 2026
- All accounts
- global
- Fetched, browser-rendered
- 49193261…CEE4
How we read it: The privacy policy lists "to train our artificial intelligence/machine learning models" among its purposes, and states two mitigations: content is disassociated from account information first, and the initial models were trained on public data rather than user chats. What makes this D is where the exit lives — the policy's own "Your Choices" section offers deletion and an advertising opt-out but no training opt-out, and the right quoted here appears only in the Regional Privacy Disclosures, pointing at a FAQ article that 403s this monitor. Note also one limit on deletion: a Character you made and shared stays live after you delete your account.
opt out, opt-out, training, FAQ, by default, default, control
What the document does offer
- "With specific regard to the processing of your personal data for the purpose of training and improving our AI models, you have the right to opt out at any time"
- the mechanism, by cross-reference: "You may exercise this right by following the instructions provided in this FAQ article"
- "You have control over your personal data and can decide whether such data can be processed for these purposes"
- "If you would like to opt out of ‘sales’ or targeted advertising, please refer to the Rights You May Have section" — advertising, separate
- "default" appears zero times in the capture
Why none of it applies: The right is stated unusually plainly and is specific to model training rather than borrowed from a general rights list. What the policy does not carry is the mechanism — that is deferred to an FAQ article, so a reader learns the right exists and must leave the document to use it. Nor does anything state the default. D rather than C on both counts.
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.
- Self-serve toggle
- Not stated
- 49 / 100
- Yes, stated
- Unknown
How to opt out of Character.AI training
Plan by plan, from what Character.AI'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
Yes. Character.AI has a self-serve setting.
- Log in and open Settings.
- Select "Data & Privacy".
- De-select "Improve the Model for Everyone".
Character.AI writes these steps for users in the EEA and UK. New content only once you've opted out; feedback you give, and use for search, recommendations and safety classifiers, carry on.
- Select "Account".
- Select "Manage Account & Data".
- Select "Model Improvement" and de-select "Improve the Model for Everyone".
What this policy has done since we started watching
1 change detected since 14 Jul 2026. Each one is a diff between two archived copies, and each links to the passage that moved.
Vendor response
We wrote to Character.AI on 24 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