Trust & Safety
Content Moderation
কনটেন্ট মডারেশন
The exact workflow every report passes through, who decides, and how the decision is recorded.
The workflow
- 1. Report — filed by a member with a category, description and optional evidence link.
- 2. Automated risk check — a lexicon pass plus AI-assisted triage produce a risk score. Advisory only.
- 3. Quarantine — high-risk items are hidden from public view before any human sees them.
- 4. Human review — a moderator reads the case and the preserved evidence.
- 5. Action — warn, remove, restrict, mute, suspend, ban, or dismiss.
- 6. Appeal — the member may contest the decision.
- 7. Audit history — every stage is written to an immutable log.
Limits on automation
- AI may raise a flag. It may never issue a verdict, close a case or restrict an account.
- A dismissed report leaves no mark on the reported member's record.
- Moderators cannot delete or edit safety log entries — not even administrators.
Tools you have
- Report
- Track your report
- Appeal
Actions the safety desk can take
- Moderation queue
- Risk triage
- Case history
- Appeal decisions
Report something under this policy
Reports filed here cover Spam, Deceptive misinformation, Dangerous or illegal content, Pornography or sexualised content. Sign in first so we can notify you of the outcome and so you can appeal if you disagree with it.