How It Works
🏛️ Town Hall Format
Behind Charter Schools is a public discussion platform focused on charter education. Everyone sees everything - no algorithms hiding posts, no echo chambers. Just honest, open discussion.
📝 Creating Discussions
- Any registered user can start a new discussion thread
- Threads can be scoped to a specific state or kept national
- All threads are automatically fact-checked by AI
💬 Commenting
- Add your perspective to any discussion
- Comments can be fact-checked when 3+ users request verification
- Rate-limited to prevent spam (30 comments per hour)
- Fact-check requests help maintain accuracy in discussions
👍 Reactions
- Like - Support the post
- Dislike - Disagree with the post
- Fact Check - Request AI verification
✅ AI Fact-Checking
Both threads and comments can be fact-checked using AI powered by 
Curio.chat.
How Fact-Checking Works
There are two ways fact-checking is triggered:
When you create a new thread, it's automatically queued for fact-checking. No votes needed - every new discussion gets verified.
When 3+ users click the "Fact Check" button on any comment, it's automatically queued for AI verification. This allows the community to flag important claims made in discussions for review.
The AI Process:
- Extract Claims: AI identifies verifiable factual claims from the content
- Search Evidence: Uses Brave Search to find supporting or contradicting sources
- Source Scoring: Weights sources by credibility (.gov=90%, .edu=85%, major news=80%)
- Verify Claims: Compares each claim against the evidence found
- Generate Verdict: Determines if content is Verified, Disputed, Partially True, or Unverifiable
- Display Results: Shows findings in a "Powered by Curio.chat" bubble with confidence score
Processing Times:
- Automatic thread checks: Processed immediately when thread is created
- Comment fact-checks: Processed every 10 minutes when 3+ votes are reached
- Background processing: Runs continuously via cron jobs
Understanding Fact-Check Results
All major claims are supported by credible sources. The post is factually accurate based on available evidence.
Major claims contradict credible sources. The post contains significant factual errors or misinformation.
Some claims are accurate, others are not. The post mixes factual information with errors, exaggerations, or missing context.
Claims cannot be confirmed or denied with available sources. This may be due to opinions, predictions, or lack of public information.
What Do Confidence Scores Mean?
The percentage (e.g., "78% confidence") indicates how certain the AI is about its verdict:
- 90-100%: Very high confidence - Strong evidence from multiple authoritative sources
- 75-89%: High confidence - Good evidence, but some ambiguity or conflicting sources
- 60-74%: Moderate confidence - Limited evidence or significant uncertainty
- Below 60%: Low confidence - Insufficient evidence or highly contested claims
💡 Example: A result showing "PARTIALLY TRUE (78% confidence)" means the AI found that some claims are accurate and others are not, and it's fairly confident (78%) in this mixed assessment based on the sources it found.
Note: Fact-checks are AI-generated and should be used as a starting point for further research. Always review the sources provided and think critically about the information.
🔔 Notifications
- Get notified when new discussions are posted
- Receive updates when someone comments on your thread
- Know when fact-checks complete
- Control email notifications in your profile