Understanding the mechanisms of induced nuclear fission for a broad range of neutron energies could help resolve fundamental science issues, such as the formation of elements in the universe, but could have also a large impact on societal applications in energy production or nuclear waste management. The goal of this paper is to set up the foundations of a microscopic theory to study the static aspects of induced fission as a function of the excitation energy of the incident neutron, from thermal to fast neutrons. To account for the high excitation energy of the compound nucleus, we employ a statistical approach based on finite-temperature nuclear density functional theory with Skyrme energy densities, which we benchmark on the 239 Pu(n,f) reaction. We compute the evolution of the least-energy fission pathway across multidimensional potential energy surfaces with up to five collective variables as a function of the nuclear temperature, and predict the evolution of both the inner and outer fission barriers as a function of the excitation energy of the compound nucleus. We show that the coupling to the continuum induced by the finite temperature is negligible in the range of neutron energies relevant for many applications of neutron-induced fission. We prove that the concept of quantum localization introduced recently can be extended to T > 0, and we apply the method to study the interaction energy and total kinetic energy of fission fragments as a function of the temperature for the most probable fission. While large uncertainties in theoretical modeling remain, we conclude that finite-temperature nuclear density functional may provide a useful framework to obtain accurate predictions of fission fragment properties.