Ensure the AI system’s language and behaviors do not reinforce undesirable and unfair stereotypes and biases.
Personalize the user’s experience by learning from their actions over time.
Limit disruptive changes when updating and adapting the AI system’s behaviors.