The panchromatic, broad-band, spectral energy distribution (SED) of galaxies is usually modelled by combining together the theoretical spectra of its emission components: stars in the optical/near-infrared, and thermal emission by dust -heated by the stellar radiation field- in the infrared. SED fitting codes such as MAGPHYS and CIGALE are capable to automatically fit observed multiwavelength data of galaxies, providing a set of galactic properties as a result. The situation gets somehow complicated when Active Galaxies (both local, low-luminosity Seyferts, and the bright QSOs) are considered. Very often, in fact, their observed near- and mid-infrared (NIR and MIR, respectively) SED is dominated by the emission of hot dust located close to the supermassive, active black hole which powers the bulk of their luminosity. Hence, a third component must be added to the set of theoretical SEDs: that of the molecular torus which surrounds the disk of gas accreting onto the supermassive black hole. The standard way to do it, is to simply add such models to the observed SED, until the MIR gap is filled. This implicitly assumes that the AGN has no influence whatsoever on the dust properties on scales larger than that of the torus (~few pc). I am investigating whether this assumption is valid, in which cases, and under which circumstances the AGN provides a non negligible contribution to the interstellar radiation field heating the diffuse dust in galaxies. This is accomplished by means of radiative transfer models which take into account the most relevant characteristics of the problem: the relative dust-stars distribution and the very wide range of spatial scales involved.