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.
Immediately update or convey how user actions will impact future behaviors of the AI system.