Critique becomes risky when it is implicit. In a historical perspective and following the etymological roots, the paper conceptualizes different modes of critique and makes their epistemological foundations explicit. The idea of a critique of data was for the first time presented by Agentur Bilwet and Frank Hartmann in the 1990s. Following these traces, the article explores how Michel Foucaults genealogical model of critique can be applied to the objects of data-critique. Consequently, it asks how critique itself can be accomplished by data, how we can judge on the given, and finally, what all this means for media studies.