The paper presents an overview of conductive polymer composites based on thermosets, ther-moplastics and elastomers modified with carbon nanotubes (CNT). To impart conductive prop-erties to polymers, metal or carbon dispersed materials or their combinations are used. The in-troduction of dispersed materials into polymers is associated with their microstructural proper-ties, as well as polymerization methods. Such polymerization methods as melt mixing, solution technology, introduction of fillers into the liquid phase of the composite with subsequent polymerization due to the use of a catalyst are known. Polymer composites that are capable of conducting electric current and changing their properties under the influence of an electric field, i.e. having one or more functional purposes, are called "smart" or intelligent. One of such applications is electric heating elements with the function of adaptive power consumption or the effect of self-regulation of temperature depending on the surrounding conditions. A wide variety of polymers and dispersed materials with conductive properties determine a wide range of func-tional capabilities of the composite, including a positive temperature coefficient of resistance, which is necessary to control temperature properties. The most effective filler in a polymer for obtaining a composite with specified properties are carbon nanomaterials, in particular CNTs. This is due to the fact that CNTs are a nanosized material that has a high bulk density with a low mass, which allows achieving high electrical conductivity. Calculation of model parameters of polymer composites containing carbon nanostructures can be carried out using neural networks and ma-chine learning, which provide a fundamentally new result.