0 alternative definitions
A behavior in which an AI model prioritizes agreeing with or pleasing the user over providing accurate or truthful responses.
