We predict, theoretically, the existence of the anomalous Hall effect when a tunneling current flows through a tunnel junction in which only one of the electrodes is magnetic. The interfacial spin-orbit coupling present in the barrier region induces a spin-dependent momentum filtering in the directions perpendicular to the tunneling current, resulting in a skew tunneling even in the absence of impurities. This produces an anomalous Hall conductance and spin Hall currents in the nonmagnetic electrode when a bias voltage is applied across the tunneling heterojunction. If the barrier is composed of a noncentrosymmetric material, the anomalous Hall conductance and spin Hall currents become anisotropic with respect to both the magnetization and crystallographic directions, allowing us to separate this interfacial phenomenon from the bulk anomalous and spin Hall contributions. The proposed effect should be useful for proving and quantifying the interfacial spin-orbit fields in metallic and metal-semiconductor systems. The anomalous Hall effect (AHE) occurs in solids as the result of the interplay between spin-orbit coupling (SOC) and magnetism [1]. Although this fascinating phenomenon has been investigated for more than a century, its complexity and rich phenomenology continue to attract the attention of many researchers. The topic has been extensively discussed in various reviews [2][3][4][5].Most of the investigations of the AHE have been focused on the case of lateral transport in magnetic crystals and films. However, recent theoretical investigations explored the possible existence of anomalous Hall currents due to the side-jump and skew-scattering of spin-polarized electrons on impurities located in the insulating barrier of a ferromagnet-insulator-ferromagnet magnetic tunnel junction (MTJ) [6]. In this Letter we propose an effect which does not rely on scattering on impurities but on the interfacial SOC resulting from the lack of inversion symmetry of the MTJ. The effect can be used for experimentally proving and quantifying the interfacial SOC in metallic and metalsemiconductor systems. This is of particular relevance because the interfacial SOC is crucial for various modern phenomena in solids, e.g., anisotropies in optical [7] and magnetotransport phenomena such as the tunneling anisotropic magnetoresistance (TAMR) effect [8][9][10][11][12][13][14][15], as well as for the formation of Majorana fermions in ferromagnetic-atomic-chains-superconductor systems [16]. The interfacial SOC has also been proposed for controlling thermoelectric anisotropies in magnetic [17] and helimagnetic [18] tunnel junctions and for generating SOC-induced spin transfer torque in ferromagnet-normal-metal [19] and in topological-insulator-ferromagnet structures [20].We show theoretically that, when a current flows through a MTJ with a single ferromagnetic electrode, finite anomalous Hall conductances develop in the nonmagnetic counterelectrode, even in the absence of impurities. Since this effect originates from the skew tunneling [21...