Intelligent GIS store the accumulated experience of observation and decision-making in spatial situations. Representations and judgments regarding previously observed situations are significantly influenced by the observer’s perception of space and time. Consequently, cognitive biases are inevitable, leading to a reduction in the reliability of knowledge. This paper addresses the problem of detecting and imprecisely assessing the degree of knowledge reliability in spatial situations affected by cognitive biases. The focus is on cognitive distortions arising from incorrect perception of the geometric and semantic properties of objects and relationships in the real world. A precedent of a spatial situation is cartographically represented within a GIS. The approach to detecting cognitive biases is based on the premise that situations sharing a common context can be cartographically transformed into one another. The algorithm constructs a cartographic representation of the analyzed situation within a given spatial domain. It is assumed that the reliability level of a transformed situation can serve as a measure of cognitive distortions. This enables a fuzzy characterization of knowledge about spatial situations. Subsequent spatial analysis and decision-making must take into account the degree to which a situation belongs to knowledge that is free from distortions. The concept of a spatial situation image is introduced. This image consists of two components: the first is a precise precedent, while the second is a fuzzy set of generalizations concerning permissible, impermissible, and possible transformations of the precedent within a given context. A procedure for transformation within a specified spatial domain is described, along with reliability assessments of the transformation results. It is proposed that the degree of cognitive biases can be evaluated through the reliability levels of transformed images stored in the GIS knowledge base.