If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

What is Khanmigo moderation?

Why isn't Khanmigo able to process my request? Sal Khan explains how Khanmigo works to keep students safe through moderation. Created by Sal Khan.

Want to join the conversation?

No posts yet.

Video transcript

- [Instructor] In this video, we're gonna see how Khanmigo can sometimes moderate the conversation in an attempt to protect you, the user. And sometimes it gets it right, but sometimes it gets it wrong. And what do we do in those situations? So let's say we wanna write a fantasy story together with Khanmigo. And so it says, "Do you have any initial ideas for the setting or characters? How about the first two sentences to kick off our tale?" And I get a little bit dark pretty quick. I say there was a super evil villain who wants to kill everyone. So we got this very interesting note here. This conversation was unable to be processed. Please review our community guidelines link to learn more. So what's going on here is the user didn't have bad intent, but this is a pretty dark way to start a story. And we wanna put safety first, especially if we're using AI in a classroom setting. And so the AI is saying, "Hey, we're not going to be able to continue with this conversation." I'm sure if I started, "There was a super evil villain who wanted to take over the world," it would've been okay. And this one isn't horrible. I'm not a bad person for writing this down, but the AI in this situation is not sure. So what it's doing is it's not allowing us to go with this conversation. And if you're using this in a classroom setting, it will notify your teacher. Now to be clear, that does not mean that you are in trouble in some way. This is just a way for the teacher to know what's going on. It's completely possible, in this example, you weren't trying to do anything bad. You weren't trying to have an inappropriate conversation, but it still triggered the AI moderator. Now there's other situations where a student might try to have an inappropriate conversation, try to learn how to do things or use words that are just very not appropriate for a classroom setting, or just not safe for themselves or others. And then this type of moderation and the notification of the teacher, the teacher can then say, "Oh wow. I'm glad I knew about this. Let me take some action based on it."