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.
Personalize the user’s experience by learning from their actions over time.