Transparency
Last updated: 20 June 2026
We believe the people who use Meowmbai deserve to know how the platform operates, how content is moderated, and how we handle external requests. This page is updated with each transparency report.
Moderation (May 2026 - Launch period)
| Metric | Count |
|---|---|
| Total uploads reviewed | 50 |
| Auto-approved (≥80% confidence) | 38 |
| Sent to human review | 12 |
| Human-approved | 10 |
| Rejected | 2 |
| Reason: No cat detected | 1 |
| Reason: Identifiable human | 1 |
| User appeals received | 0 |
| Average review time | ~4 hours |
User reports
| Report type | Count | Resolved |
|---|---|---|
| Incorrect location | 0 | - |
| Copyright claim | 0 | - |
| Privacy concern | 0 | - |
| Inappropriate content | 0 | - |
| Spam / duplicate | 0 | - |
Law enforcement requests
In the launch period, Meowmbai received zero law enforcement data requests.
Government removal requests
Zero government content removal requests received.
How we make moderation decisions
Every upload is first evaluated by an AI system that checks for cat presence and estimates confidence. High-confidence results (≥80%) are auto-approved. Lower confidence and edge cases are reviewed by a human moderator. We do not outsource moderation. Moderators follow documented guidelines and can be appealed.
Location fuzzing
All map markers are offset by a random 50–150 metre radius from the contributor's pin. Exact coordinates are never exposed publicly. This is implemented at the database level and cannot be bypassed through the public API.
Next report
We will publish our next transparency report covering Q3 2026 activity in October 2026.