Glossary
Content Moderation
Content moderation is the practice of reviewing, filtering or removing user-generated content in a messaging environment to enforce community guidelines, legal obligations or platform safety standards.
General definition
Content moderation in chat covers the policies, tools and workflows used to ensure messages, images, files and other user-generated content meet acceptable-use rules. It applies to consumer social products, enterprise tools and any platform where users can communicate with one another.
- Automated filtering: rule-based or machine-learning classifiers that flag or block content matching predefined categories (profanity, spam, hate speech, explicit material)
- Human review: moderators who investigate flagged items and apply judgement
- User-level controls: blocking, muting and reporting between individual users
- Admin controls: delete message, suspend user, set channel-level permissions
Moderation can be pre-publish (content is held for review before delivery), post-publish (content is visible immediately but can be removed after review) or a hybrid. For regulated industries such as healthcare, content moderation also covers data-loss prevention and ensuring that sensitive information is not shared outside permitted channels.
In the Ethora ecosystem
Ethora provides built-in moderation controls through the Chat SDK and admin dashboard: administrators can delete any message, suspend or ban users, set room-level permissions and review flagged content. These controls apply across all room types, including group rooms, channels and one-to-one conversations.
For AI-powered moderation, the AI SDK lets you attach a content-screening agent to a room or channel. The agent evaluates incoming messages against a policy and can block, flag or auto-reply before the message is stored or delivered. Combined with self-hosted deployment, this means the moderation logic and the classified content never leave your infrastructure.
