Help the user understand what the AI system is capable of doing.
Ensure the AI system’s language and behaviors do not reinforce undesirable and unfair stereotypes and biases.
Enable the user to access an explanation of why the AI system behaved as it did.