We direct the AI models to moderate content so as to decline certain requests based on age, a school based setting, and inappropriate content as layers on top of the base models we use that have their own safeguards in place as well.
We’re also actively improving content moderation and benefit from being the most used AI platform for education with over 5 million users who give us feedback when they discover an issue. Find a bug or content that you think MagicSchool shouldn’t display? E-mail us at hey@magicschool.ai and we’ll address it.