AI moderation

AI moderation reads messages in the channels you choose and flags toxicity, scams and phishing with more context than a word list can.

What it detects

• Toxicity: harassment and abusive language
• Scams: common scam patterns such as fake giveaways
• Phishing: links and messages that imitate trusted services

Turn it on

1. Open AI moderation in the dashboard.
2. Choose the channels to cover.
3. Set a sensitivity from relaxed to strict.
4. Choose what happens to flagged messages.
5. Save changes.

Review queue

Borderline cases go to a review queue instead of being punished automatically. A moderator can confirm or dismiss each item, and the result is recorded as a case.

Privacy

Message content is only processed for the channels you enable, and is not stored by AI moderation unless a message is flagged for review. See Your data.

Commands