These terms describe instances where AI models generate responses that contain false or misleading information presented as fact. in AI, it is associated with erroneously constructed responses (confabulation)