The question of validity is indeed crucial; one must always anticipate the inherent weaknesses of any statistical comparison. Before asserting a statistical similarity, it is imperative to ensure that the initial conditions of non-fraud detection are met for both data sets, because an absence of divergence can also mask systemic problems if the context is not rigorously controlled. Without this verification, the Kolmogorov-Smirnov test, despite its power, could mask risks rather than reveal them, such as asymmetric sector regulation that would distort the distributions.