# Kuan Yin — Frequently Asked Questions

> User guide 32.16.2.8.

**What is Kuan Yin?** It's the part of the platform that helps keep
conversations kind and people safe. Instead of just punishing rule-breaking, it
offers gentle pauses before harm happens, protects people who are targeted, and
helps those who slip up find their way back.

**Is it reading my private messages?** It analyses content in the moment to make
a safety decision, using a one-way fingerprint, and then forgets it. It doesn't
keep your messages to mine later. See the
[Privacy controls guide](07-privacy-controls.md).

**Why did I get a pause before posting?** Your message looked like it might
cause harm. The pause is optional — you can skip it and post anyway, with no
hidden penalty. See [Mindful moments](01-mindful-friction.md).

**Did I lose points for skipping a prompt?** No. Skipping a prompt carries no
secret penalty.

**What is "merit" / "karma"?** A record of your positive contributions that
builds trust and unlocks privileges. It's private by default and can't be bought
or transferred. See the [Merit system](03-merit-system.md).

**I was blocked/timed out and I think it's wrong. What do I do?** Appeal it — a
human reviews significant actions. See [Appeals](06-appeal-process.md).

**I'm a creator getting harassed. Can you help?** Yes — that's what the shield
is for. See [Creator protection](05-performer-protection.md).

**Someone shared a fake/intimate image of me. What now?** Report it immediately.
Kuan Yin works to detect, remove (on strict timelines), block re-uploads,
preserve evidence, and connect you with support.

**I'm having a really hard time / thinking about hurting myself.** You're not
alone, and help is available right now. Kuan Yin will connect you to a real
person and to crisis resources for your region. Please reach out.

**Can I turn the predictive features off?** Yes — you can opt out of predictive
features and adjust how much friction you see.

**Does a human ever make these decisions?** Yes. Irreversible actions like bans
are never finalised by automation alone — a person reviews them.

**Where can I learn what a term means?** See the [Glossary](09-glossary.md).
