Help the user understand how often the AI system may make mistakes.
Display information relevant to the user’s current task and environment.
Ensure the AI system’s language and behaviors do not reinforce undesirable and unfair stereotypes and biases.