Does Snap train on your data?
Snap says it may use what you post publicly, on Spotlight, Public Stories and the Snap Map, to train its generative models unless you switch it off in Settings. Private snaps and chats: not stated.
A self-serve opt-out exists
"Snap may use the images, videos, audio, and text content that you have posted publicly on Snapchat (such as posts to Spotlight, Public Stories, and the Snap Map) to develop, train, and improve Snap's generative AI content creation models. This may include a combination of automated and human review of this public content."
- 21 Sept 2026
- 5 Oct 2026
- Public content (Spotlight, Public Stories, Snap Map)
- global
- Fetched from the company
- DC6855AA…1674
How we read it: Resolved from UNCLEAR on 10 September 2026, not by the privacy policy but by Snap's support article "Generative AI on Snapchat", which says in one paragraph what the policy never does: public content trains Snap's generative models, people may review it, and a setting under Privacy Control turns it off. Two limits in Snap's own words: the opt-out works "going forward" and "does not affect training that has already taken place", and it does not cover you appearing in someone else's public content. The article names the switch and its location but never says which way it ships. Private snaps, chats and My AI conversations are the second tier, still unresolved.
opt out, Settings, Privacy Control, default, going forward, human review
What the document does offer
- "In Settings, you can opt out of allowing Snap to use your public content for these purposes."
- the steps: Settings, then "Under 'Privacy Control,' tap 'Manage My Information'", then the toggle
- "opting out does not affect training that has already taken place"
- "Snap may still process and use information about you if you appear in public content posted on Snapchat by another user."
- "default": 0 matches
Why none of it applies: The control is named and located: Settings, Privacy Control, Manage My Information, a toggle. The article states no default; "default" does not appear, and the training sentence is "Snap may use" with the opt-out offered after it, which is how a switch that ships on is usually described. What the opt-out does not reach is stated by Snap: training already done, other uses under the privacy policy, and your appearance in other people's public content. Nothing in the article or the policy extends any of this to private content.
The support article that grades the tier above covers public content only. For private snaps, chats between people and conversations with My AI, the privacy policy is what there is, and it says that Snap's "algorithms and machine learning models take into account the conversations Snapchatters are having with My AI to improve the responses" — a sentence that sits equally well with training a model on those conversations and with feeding them in as context at answer time. The policy never picks, and says nothing either way about snaps or about chats between people. Corrected 19 Aug 2026 from a reading that saw only a definition; still UNCLEAR on 10 Sept 2026, now narrowed to private content.
- Snap privacy policy, captured 2026-09-07 (https://snap.com/en-US/privacy/privacy-policy) · Snap support article "Generative AI on Snapchat", captured 2026-09-10 (public content only)
- train, training, fine-tune, machine learning, artificial intelligence, AI model, large language model
- 0 of 8 raw hits
- 10 Sept 2026
Four sentences carry the terms. One defines generative AI, one defines how generative models learn from training data, one lists the purposes Snap uses machine learning for, and one — quoted in this verdict — says those models take conversations with My AI into account to improve it. None states whether user content is used to train a model, which is the question this site asks.
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
- 53 / 100
- Not stated
- Yes
- Not established
- Not stated
- —
- —
- Not established
How to opt out of Snap AI training
Plan by plan, from what Snap'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.
Public content (Spotlight, Public Stories, Snap Map)
Yes. Snap has a self-serve setting.
- Tap the settings gear to open Settings.
- Under Privacy Control, tap Manage My Information.
- Turn off the setting for Snap using your Public Content to train generative AI models. It applies going forward only.
Private snaps, chats and My AI conversations
Not graded, so no route is pointed at. The support article that grades the tier above covers public content only. For private snaps, chats between people and conversations with My AI, the privacy policy is what there is, and it says that Snap's "algorithms and machine learning models take into account the conversations Snapchatters are having with My AI to improve the responses" — a sentence that sits equally well with training a model on those conversations and with feeding them in as context at answer time. The policy never picks, and says nothing either way about snaps or about chats between people. Corrected 19 Aug 2026 from a reading that saw only a definition; still UNCLEAR on 10 Sept 2026, now narrowed to private content.
What this policy has done since we started watching
No change detected. We've held a copy of Snap's documents since 13 Aug 2026 and re-read them since, across 4 captures of 2 documents. The text hasn't moved in a way that touches this verdict. That's a fact about the documents, and it's only knowable by having kept the earlier copies.
Vendor response
We haven't yet written to Snap about this verdict, and they haven't disputed, clarified or announced a change to any clause quoted above. When either happens it's published here, word for word.
Report a wrong clause