Flagging & Escalation System

Written By Christopher Lee

Last updated 6 months ago

Knobase includes an AI-powered Flagging & Escalation System that automatically detects and responds to prompts containing sensitive, harmful, or inappropriate content. This feature is essential for maintaining a safe and supportive learning environment, especially when students interact with AI agents independently.

🧠 Why Flagging Matters
✅ Protects students from exposure to harmful or unsafe content
✅ Alerts educators to potential well-being concerns
✅ Promotes responsible AI usage and digital citizenship
✅ Enables early intervention for at-risk students

⚙️ How It Works

  1. Automatic Detection
    When a student types a prompt, Knobase scans it for red flags using built-in safety filters and contextual analysis.

  2. Flag Categories
    Prompts are flagged if they contain:

    • Bullying or harassment

    • Harmful advice or self-harm indicators

    • Racial or political sensitivity

    • Inappropriate or unsafe language

    • Emotional distress signals

  3. Escalation Protocol
    If a prompt is flagged:

    • The system sends an alert to designated staff (e.g., counselor, safeguarding lead)

    • The flagged prompt is logged in the admin dashboard for review

    • Optional: The student may receive a supportive message or be redirected to appropriate resources

Check “Chats & Messages” section in the Dashboard

View “Flagged” Message history

Choose to “Dismiss” or “Escalate” / Option to view the entire chat

📊 Use Cases 🧑‍🎓 Students: Protected from unsafe AI interactions
🧑‍🏫 Teachers: Gain visibility into student well-being and digital behavior
🧑‍💼 Admins: Monitor flagged prompts and ensure timely intervention