# The AI model somehow says the user's answer incorrect and then says the user's answer in its message. Is there a way to fix this?