The article continues the exploration of “non-ideal” epistemology by focusing on collective epistemic attitudes such as prejudice and stereotypes. These attitudes are examined in terms of their epistemic status, distinct from moral evaluation. The article challenges the view of prejudices and stereotypes as inherentlyfalse by analyzing their functions of cognitive economy and social legitimation, first noted by W. Lippmann. The distinction between the truthfulness and correctness (validity) of a stereotype as an epistemic attitude is emphasized, where the priority is given to the procedural aspect of belief formation. The article supports a probabilistic approach to assessing stereotype accuracy, treating stereotypes as “generic statements” whose truth should be assessed empirically rather than a priori.