What we remove, what we keep

baatao is built for honest argument. This page tells you exactly what crosses the line, what doesn’t, and what happens when a report is filed. We’d rather over-explain than surprise you.

• Targeted harassment. Repeated personal attacks, doxxing, stalking, sustained brigading of one person.
• Hate. Slurs, dehumanizing speech, or content that attacks people for who they are.
• Threats of violence, sexual content involving minors, gore, or content that promotes self-harm.
• Spam, coordinated inauthentic behavior, or paid vote brigades.
• Doxxing of private people. Phone numbers, home addresses, government IDs, etc. Public figures’ public channels are different.
• Recordings without consent in jurisdictions where both-party consent applies. We require consent at the lobby step.

• Opinions you disagree with. That’s the product. Strong takes are the point.
• Losing positions. Coming second isn’t a policy violation.
• Criticism of public figures on matters of public interest, including pointed criticism of policy, business conduct, or public statements.
• Satire + parody that’s clearly identified and doesn’t cross into harassment or defamation.

We use two automatic thresholds. Both are conservative and both ratchet — they don’t auto-undo once a moderator gets eyes on:

• Auto-hide on 5 reports against the same content. A comment, chat message, or debate that crosses the threshold is hidden until a moderator reviews. The author is told. Vote integrity stays intact (we don’t wipe the row).
• Priority bump on 3 reports across the same author. Every open report against them jumps to high priority for a faster triage.

Numbers are tunable — what’s here is the current default. Both triggers leave audit trails so we can review false-positives.

We weight votes by account age, vote history, and behavioral signals — we publish a vote integrity card on every debate so you can see the trust composition without trusting our final tally.

If the fresh-account share of votes crosses 60% on a debate with at least 25 votes, we freeze voting for moderator review. The freeze is disclosed inline. Moderators can unfreeze with a recorded reason.

1. You file the report. We record it.
2. If your report crosses an auto-mod threshold, content is auto-hidden.
3. A moderator reviews. Status moves to reviewing.
4. We either remove the content (and tell the author why), restore it (and tell you why), or escalate if it’s ambiguous.
5. You see the resolution in your notifications.

If your content was removed and you think we got it wrong, email appeals@baatao.com with the debate / comment URL and a short explanation. We respond within 7 days. A second human reviews — not the same moderator who made the original call.

• Take down your own recording from the watch page. The debate, votes, transcript, and result stay.
• Redact your own transcript if you said something you’d rather not have searchable. Same page.
• Delete your account from /settings/privacy. 30-day grace; full export available on the way out.

• Sell verification.
• Sell vote weight.
• Run behavioral ads.
• Voluntarily share data with anyone.
• Hide our moderation actions from you.