Why did Khanmigo flag my conversation?
Khan Academy uses a moderation system that combines large language models and other technologies to review conversations and flag content that may be unsafe or inappropriate for a learning environment. These tools are supplemented by human review to help maintain a safe experience for all users.
If one of your conversations was flagged, that means our system detected something that might go against our Khanmigo Community Guidelines.
What’s considered inappropriate?
A good rule of thumb is:
If it wouldn’t be appropriate to say out loud in a classroom, especially with elementary students present, it probably isn’t appropriate on Khan Academy either.
Khanmigo’s moderation system may flag conversations that include:
- Profanity or inappropriate language
- Violence, threats, or references to harm — even if meant as a joke
- Sexual or other mature content
- Phrases that may trigger safety filters, including references to self-harm or self-harm ideation
- Bullying, harassment, or hate speech
Can the system make mistakes?
Yes — and we know it sometimes does.
Like many AI systems, Khanmigo’s moderation tools are continually improving. We use a combination of automated systems and human review to help strengthen accuracy over time. Because the system is designed to prioritize safety, it may occasionally flag content that is actually appropriate.
We understand this can be frustrating, especially when you’re asking a thoughtful question or seeking help with your schoolwork. That’s why we regularly review flagged cases, learn from them, and refine our processes to improve accuracy and reduce unnecessary flags.
Want to learn more?
You can read the full Khanmigo Community Guidelines here.
What happens if my child or student's Khanmigo conversation gets flagged?
If you think your message was flagged by mistake and you have questions, you can also contact our support team.