Does Meta AI train on your data?
Meta AI is trained on your chats with it and on your public posts by default, Meta says on its own AI page. You can object through your account, and future training runs will skip you.
"This includes the right to object to your interactions with features that are part of AI at Meta and the public information that you've shared on Meta Products being used to develop and improve AI at Meta."
- 6 Sept 2026
- All accounts (Meta AI in Facebook, Instagram, Messenger and the Meta AI app)
- uk
- Saved by hand — not re-read automatically
- E33FC6CB…32F7
How we read it: The page lists its training sources, among them "Your interactions with features that are part of AI at Meta" and "Public information such as public posts and comments shared on Meta Products"; says private messages are not used unless shared with an AI; and rests the processing on legitimate interests in the UK and EU. The route out sits beside the right: an objection request through your Facebook, Instagram or Meta account, confirmed by email, applied to future training runs. A form to find rather than a setting, so a D. The AI terms add that Meta may review your conversations with AIs, "automated or manual (human)". Both pages were printed to PDF by hand and read by OCR.
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
- 58 / 100
- Not stated
- Yes
How to opt out of Meta AI training
Plan by plan, from what Meta 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 (Meta AI in Facebook, Instagram, Messenger and the Meta AI app)
Yes, but not self-serve: Meta AI makes you contact support.
- Log in to the account you use with AI at Meta.
- Open facebook.com/privacy/genai/ and, under "You have rights related to how your information is used for AI at Meta", follow the link for your account to the objection form.
- Submit the objection. Meta emails a confirmation; an objection made after a training run has started applies to future runs.
What this policy has done since we started watching
The record starts here. Meta AI's 2 documents were first captured on 6 Sept 2026, and there's no earlier copy to compare them against, so this verdict has no history behind it yet. From now on every re-read is diffed against those copies, and anything that moves appears here.
Vendor response
We wrote to Meta AI 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