AI Moderation: Reason Codes

Settings > Moderation > Spam Management > Reason Codes

Reason Codes label moderation actions, telling users why their content was actioned and helping your team and AI categorize decisions consistently. They are visible in the AI Moderation Analysis modal.

System Reason Codes

System Reason Codes are built in and used automatically by AI moderation. You can edit the user-facing message but cannot delete them. 

Key codes include:

• FRAUD_IMPERSONATION - This post impersonates another person or organization.
• FRAUD_SCAM - This post appears to be a scam or fraudulent offer.
• HAM_APPROVED - This post is acceptable and has been approved.
• HARM_HARASSMENT - This post contains harassment, bullying, or abusive language.
• HARM_HATEFUL - This post contains hateful or discriminatory content.
• SPAM_PROMOTIONAL - This post contains unwanted promotional or commercial content.

Additional System codes can be viewed under Spam Management,

Custom Reason Codes

Define violation types specific to your community. Once created, AI moderation will use them to categorize matching content. Codes can be disabled at any time without deleting them.

Creating a Custom Reason Code (3 steps):

Step 1 - Category

Step 2 - Details: 

Enter a Violation Name (e.g. "advertising" → auto-generates SPAM_ADVERTISING) and a User-Facing Message (e.g. "This post is trying to advertise commercial products").

Step 3 - Review: 

Confirm the Technical Code, Display Name, Category, and User Message, then click Create Reason Code.

The technical Code ID is immutable once created. You can update the Display Name or user-facing message at any time. Deactivate codes you no longer use and historical data is preserved. You can only delete codes that have never been used.