[en] UO₂ and (U,Pu)O₂ are the most used fuels in the current nuclear reactor fleet. Under irradiation in a reactor, new chemical elements are created in these materials by the fission of uranium and plutonium nuclei. Some of these fission products, especially iodine, caesium and tellurium may chemically react with one another to form compounds that are potentially corrosive for the cladding. One way to avoid the cladding corrosion consists of preventing the formation and migration of such corrosive species within the fuel. To that end, the determination of the most stable chemical forms of iodine, caesium and tellurium in the fuel as well as the trapping and diffusion mechanisms of these elements in UO₂ and (U,Pu)O₂ is of first interest. The gain of Cr-doped UO₂ fuel on the fission gas retention has already been demonstrated experimentally. However, no data are yet available on the release of corrosive fission products in Cr-doped UO₂. With a view to simulating the behaviour of these species within Cr-doped UO₂, one has to prior identify the most favourable oxidation state and location site of Cr in UO₂, which is a much debated topic in the literature. We perform electronic structure calculations, using the Hubbard-corrected density functional theory (DFT+U) to evaluate the preferred trapping site of I, I₂, Cs and Te in UO₂ and (U,Pu)O₂ crystals, as well as the preferred oxidation state and location of Cr in UO₂. We first determine the stability of I, I₂, Cs, Te and Cr in various point defects and then compute their XANES spectra in each considered site, using the FDMNES code and the DFT+U atomic configurations. The comparison of the computed spectra with the experimental ones contributes to the identification of the chemical forms and the trapping sites of iodine, caesium, tellurium and chromium in the actinide oxides. The use of a Hubbard term (GGA+U) allows us to take into account the strong correlations of the actinide 5f electrons. To avoid the metastable states inherent to this method, we use the occupation matrix control (OMC) procedure, which also allows us to control the valences of each species in the simulation. This particular point makes our approach reliable with respect to the determination of fission products incorporation energies in the various studied defects. (authors)