Schools
Content Safety
Loclie uses a 3-layer content safety pipeline to protect school communities where minors are present.
Time to complete: 3 minutes
Overview
Every post and comment within a school group is checked before it is published. Content that fails any layer of the safety pipeline is rejected and never appears in the group. This provides enhanced protection for school environments.
The 3-Layer Safety Pipeline
Layer 1: Keyword Filter
What it does: An instant check against a list of known explicit and inappropriate terms.
- Runs first, in milliseconds
- Catches obvious explicit content immediately
- No AI processing required — fast and reliable
Layer 2: AI Moderation
What it does: OpenAI's moderation API analyses the content for harmful categories.
- Hate speech - Content targeting groups based on race, religion, gender, etc.
- Sexual content - Explicit or suggestive material
- Violence - Graphic violence or threats
- Harassment - Bullying, intimidation, or targeted abuse
- Self-harm - Content promoting self-harm or suicide
This layer catches nuanced content that keyword filters alone would miss.
Layer 3: Enhanced School Protection
What it does: Llama Guard 3 provides an additional layer of analysis specifically calibrated for environments where minors are present.
- Activated for all school groups
- Higher sensitivity thresholds than standard moderation
- Catches content that may be technically allowed elsewhere but is inappropriate for school communities
- Provides an extra safety net beyond keyword and AI moderation
Three Layers, Not One
Each layer catches different types of inappropriate content. Running all three in sequence means content must pass every check before it is published.
What Gets Checked
Every piece of user-generated content in school groups is checked:
| Content Type | Checked | Where |
|---|---|---|
| Group posts | Yes | Before appearing in the group feed |
| Post comments | Yes | Before appearing under a post |
| Post edits | Yes | Before the edit is saved |
What Happens When Content is Flagged
- The content is rejected — it is never published or stored
- The author receives a clear message explaining that their content did not pass moderation
- The author can revise and resubmit their content
- No other members ever see the flagged content
Zero Tolerance
Flagged content is never published, even temporarily. The check happens before the post or comment appears in the group.
Human Oversight
While the automated pipeline catches the vast majority of inappropriate content, human moderators provide an additional layer of oversight.
Group Moderators
- Appointed by school staff or administrators
- Can review posts within their assigned groups
- Can remove content that the automated system may have missed
- Can manage group membership
School Staff
- Staff members with Manage portal access can moderate all groups
- Can review flagged content and take action
- Can appoint or remove group moderators
How to Appoint a Moderator
- Go to the group's settings on the Manage portal
- Find the member you want to make a moderator
- Assign the Group Moderator role
- The member can now moderate posts and comments in that group
Privacy and Visibility
School groups are designed with layered privacy:
| What | Who Can See It |
|---|---|
| School detail page | Everyone (public) |
| List of groups | School members only |
| Group names | School members only |
| Group content (posts, comments) | Group members only |
| Member list within a group | Group members only |
What This Means
- Non-school-members visiting the school page cannot see any groups
- School members who have not joined a specific group can see the group name but not its content
- Group members can see posts, comments, and other members within their group
- Parents and students only see content from groups they have been approved to join
For Parents
What You Should Know
- All content your child sees in school groups has been checked by the 3-layer safety pipeline
- Group moderators (usually teachers or school staff) provide additional oversight
- Your child can only access groups they have been approved to join
- If you have concerns about content, contact the school administrator or group moderator
Reporting Concerns
If you see content that you believe is inappropriate:
- Contact the group moderator or school administrator directly
- They can review and remove content from the Manage portal
- For urgent concerns, use the Contact Support page
Frequently Asked Questions
Can students post without moderation?
No. Every post and comment is checked by the 3-layer safety pipeline before it appears. There is no way to bypass the automated checks.
What if the automated system misses something?
Group moderators and school staff can manually review and remove any content. The combination of automated and human moderation provides comprehensive coverage.
Is the safety pipeline only for schools?
The keyword filter and AI moderation layers (Layers 1 and 2) run on all group content across Loclie. The enhanced school protection layer (Layer 3, Llama Guard 3) is activated specifically for school groups where minors are present.
Can I turn off moderation for my school's groups?
No. Content moderation is always active for school groups and cannot be disabled. This is by design to protect all members, especially minors.
Next Steps
Learn more about schools on Loclie:
- What is a School? - Overview of the Schools entity
- Creating a School - Register your school
- Managing Your School - Full management guide
Need help?

