Skip to main content

Moderation

Stream Chat has support for several moderation actions, such as flagging, muting, banning and shadow banning. You can find more details about the available low-level client moderation tools on this page.

You can moderate flagged messages and users through Stream's moderation dashboard.

Moderation Categories​

For each channel type, you can enable AI moderation, consisting of 4 moderation categories:

  • Semantic Filters
  • Commercial Spam
  • Platform Circumvention
  • Blocklists

Dashboard moderation categories

For each moderation category, you can configure an action that will be performed after a message falls into that category.

The available actions are:

  • Flag
  • Block
  • Bounce
  • Bounce then flag
  • Bounce then block

iOS SDK Handling​

In the iOS SDK, you can check if a message was moderated, by checking the moderationDetails value in the ChatMessage, which is nil for regular messages.

If a message was moderated, the moderationDetails will have the following values:

  • originalText - the original text of the message
  • action - the action that was performed (flag, bounce or block).

You can also check if a message was bounced, by checking the isBounced boolean property.

UI Components handling​

The UIKit and SwiftUI components have default handling of bounced messages. When a message is bounced, there's a visual indicator on the message view to reflect that.

Bounced message

When you long press on a message, you can perform several default actions on it, such as editing and deleting it, as well as trying to resend it.

Did you find this page helpful?