Make it easy to edit, refine, or recover when the AI system is wrong.
Enable the user to access an explanation of why the AI system behaved as it did.
Enable the user to provide feedback indicating their preferences during regular interaction with the AI system.