Last updated: 17 November 2026
In short
- We use two automated tools: a word filter on text and a nudity check on images. Both run on our own servers.
- They only flag items for review. Flagged items stay published until a person on our team looks at them.
- A person decides every action: keep the item, remove it, warn the person, or suspend the account. Nothing is removed and no strike is given automatically.
- Videos are not checked automatically yet.
- You can appeal any decision.
1. About this notice
This notice is part of the Terms of Service and the Community Guidelines. It explains how Privé, operated by POLYMATH COGNITIVE, finds and handles Content that may break our rules or the law, as the IT Rules expect.
2. How Content comes to our attention
- Reports from Users: a post report (nudity, violence, harassment, scam, impersonation, underage or other), or a report and block of a person.
- Complaints by email, including intimate-imagery and copyright complaints.
- Orders from courts and government agencies.
- Automated tools (section 3).
- Our team's own review.
3. The automated tools
| Tool | What it checks | What it looks for | What it does |
|---|---|---|---|
| Word filter | Text: chat messages, post captions, live comments, names and bios, service requests, wishlists, service listings and call packages | Words on our list (abuse, sexual solicitation and others), and phone numbers, payment IDs, links or other contact details that may be used to move Fans or payments off Privé | Adds the item to our review queue. The text is sent and published as normal |
| Nudity check | Images: photos in posts and chat, and the still image of a video where there is one | Nudity and sexual content, scored by an image-classification model | Adds the item to our review queue. The image stays visible |
- Both tools run on our own servers. Your Content is not sent to an outside company for these checks.
- The tools make mistakes in both directions: they miss some things and flag some harmless ones. That is why a person reviews everything they flag.
- The nudity check estimates nudity, not age or identity. It does not recognise faces.
- Videos themselves are not checked yet. Voice notes and calls are not checked.
- Calls and lives are not recorded or scanned.
4. Human review
A trained member of our team reviews each report and each flagged item. They look at the Content, its context, the account's history and the harm involved. They may:
- mark it as fine;
- remove the Content (it is hidden and kept as evidence);
- record a warning or strike (a record that does not restrict the account by itself);
- email the person about what needs fixing;
- suspend or close the account, for serious or repeated breaches;
- report to the authorities where the law requires.
To review a report about a private chat, staff must first type a reason, and every access is logged (see the Privacy Policy §6).
5. Telling you
- When we remove your Content or record a warning or strike, we tell you where it is safe and the law requires it, with the reason and how to appeal.
- If an automated tool flagged the item, we say so when you ask, or in the notice where we can.
- We do not tell a reported person who reported them.
6. Appeals
- A warning or strike you were told of: tap Appeal in the App within 7 days.
- Anything else, including removed Content and suspensions: write to the Grievance Officer within 30 days. We decide within 7 days. See the Grievance Redressal Policy, which also explains the appeal to the Grievance Appellate Committee.
7. Timelines
Legal deadlines (2 hours for intimate imagery, 3 hours for orders, 36 hours for other removal requests) are in the Grievance Redressal Policy. Other flagged items are reviewed as soon as we can, most urgent first.
8. Reminders
At least once every 3 months we remind everyone, in the App or by email, of our rules, what happens when they are broken, and that automated tools are used.