We will look at the current state of the AI ​​and the need to open the black box to interpret the AI’s output. Understanding how and why means maximizing performance, making the best decision, and removing bias and flaws. It is the foundation of an ethical and accountable AI. It is necessary in critical processes and B2C, but also for managers who mistrust AI once they don’t understand it. The challenge is to achieve a democratic Explainable AI, available to all companies, entities and stakeholders.