Alignment Problem

The challenge of ensuring that the goals of an AI system are aligned with human goals and preferences to prevent risks from superintelligent AI.