AI-assisted community moderation
Jatra AI
Spot the spam. Keep the judgement human.
Screen eligible public posts and replies for spam and promotional links. Give the AI your community’s context. Keep approval decisions with your moderators.
01 / Content screening
AI Spam & Link Screening
Screen contributions during publication. Hold likely spam for review before it reaches the public conversation.
- Public-content screening
Eligible posts, replies and edits are assessed during publication. Scheduled posts are checked when they are due to publish.
- Promotional-link detection
A separate check assesses unrelated commercial links. Useful external references are allowed in the screening instructions.
- Pending-review submissions
Flagged submissions stay hidden pending review. If an edit needs review, the previously approved version remains visible.
02 / Screening context
Community & Conversation Context
A recommendation can be useful in one community and spam in another. Configure the topics and promotion rules that guide your screening.
- Community moderation context
- Set context in moderation settings. If you leave it blank, screening uses your community description.
- Thread context
- Reply checks include the thread title and opening post, so the conversation informs the assessment.
- Promotion policy context
- Explain when recommendations, offers and commercial links are appropriate for your members.
03 / Human review
Human Moderation Controls
Read the submission, member details and AI signals together. Your moderators decide whether to approve or reject it.
- Spam-likelihood estimates
- Review the spam-likelihood estimate and promotional-link warning alongside the actual content.
- Approve & reject actions
- Moderators approve or reject submissions. Banning a member remains an explicit human action.
- Screening failure handling
- If a configured screening provider fails during a check, the submission is held for review without an invented score.
04 / Screening scope
Screening Eligibility
Checks focus on eligible public content from newer members and members with pending or rejected review history.
- Private-channel exemption
- Private channels bypass AI screening. The public-content screening flow skips their submissions.
- Staff exemption
- Privileged staff submissions are exempt from this screening flow. Wider moderation and access controls still apply.
- Established-member exemption
- Established members can bypass screening unless they have pending or rejected review history. Checks depend on configuration and eligibility.
Good to know
Jatra AI Questions
Does Jatra AI write posts or answer member questions?
The current feature focuses on spam and promotional-link screening. AI writing tools, summaries and a member chatbot are not included.
Does it screen every contribution?
No. Screening must be configured and applies to eligible public content. Private channels, privileged staff and eligible established members bypass these checks.
Does it automatically ban members?
No. Suspicious submissions are held for moderator review. A member ban requires a human action.
Can AI screening make mistakes?
Yes. A screening signal is an assessment, not proof. Moderators can read the submission and its context before approving or rejecting it.
Jatra AI
Explore AI Spam Moderation
Try Jatra, or book a call to walk through screening and moderator review.