Make it easy to dismiss or ignore undesired AI system services.
Make it easy to edit, refine, or recover when the AI system is wrong.
Limit disruptive changes when updating and adapting the AI system’s behaviors.