The development of lightweight materials presents challenges to constitutive modeling and numerical analysis of lightweight components. The hardening of lightweight materials varies more under different stress states than with inherent anisotropy. Although anisotropy is an intrinsic property of rolled sheets, accurate numerical analysis of lightweight components necessitates precise modeling of complex hardening under different loading conditions and anisotropy. This study characterizes the yield evolution of 5182-O aluminum alloy and employs crystal plasticity simulations to understand its plastic deformation characteristics. The mechanical properties of the 5182-O aluminum alloy were examined under different complex stress states through uniaxial tensile, plane-strain tensile, and shear experiments. Initially, the hardening behavior was accurately calibrated using inverse engineering, and plastic deformation characteristics were described analytically using the pDrucker yield equation. The pDrucker yield function was then extended to an analytical anisotropic form using an improved linear transformation tensor. The parameters of the yield function can be analyzed to model differential hardening across various stress states and anisotropic hardening along different loading directions. In addition, the evolution of voids under different stress states and grain orientations was analyzed using crystal plasticity finite element simulations combined with representative volume element (RVE) modeling. Void growth in polycrystalline materials strongly depends on the surrounding microstructure, such as grain morphology and crystallographic orientation. The RVE of single and polycrystalline aggregates containing voids was constructed using a three-dimensional Voronoi mosaic. Crystal plasticity finite element simulations were conducted to perform several simulation experiments with different combinations of grain morphology and crystallographic orientation. Results demonstrated that the anisotropic strength difference of the 5182-O aluminum alloy was < 1%, whereas the maximum strength difference between different stress states was approximately 8%, highlighting the importance of accurately modeling hardening differences due to anisotropy and stress state. The comparison of the calibrated pDrucker yield function with the experimental values under the uncorrelated flow criterion demonstrated relatively high prediction accuracy for different loading directions. The proposed yield function accurately characterized the differential and anisotropic hardening of the 5182-O aluminum alloy under various stress states. Crystal plasticity simulations revealed a strong correlation between stress triaxiality and grain orientation with evolution based on cumulative plastic slip and normalized void volume fraction.