In this study, the corrosion behavior of a commercial Al-Mg-Si alloy was systematically investigated in sodium hydroxide (NaOH) and potassium hydroxide (KOH) electrolytes over a concentration range of 1–10 M using combined electrochemical experiments and data-driven modeling. Open-circuit voltage, potentiodynamic polarization, and electrochemical impedance spectroscopy were used to quantify corrosion parameters, while X-ray diffraction assessed structural stability. The results reveal distinct corrosion regimes, with anodic activation at low-to-moderate alkalinity and transport-limited behavior at high hydroxide concentration. At equivalent molarities, KOH exhibits more negative corrosion potential (Ecorr) and higher corrosion current density (Icorr) than NaOH, reflecting cation-mediated interfacial effects rather than hydroxide concentration alone. No crystalline corrosion products were detected, indicating continuous surface dissolution with predominantly amorphous surface films. An artificial neural network optimized using a scalar genetic algorithm and a non-dominated sorting genetic algorithm (NSGA-II) achieved near-perfect predictive accuracy for Ecorr and Icorr, with R2 values exceeding 0.999. Multi-objective optimization using NSGA-II identified a robust trade-off between corrosion thermodynamics and kinetics, with an optimal condition at approximately 1.35 M KOH (Ecorr = −1.6199 V, Icorr = 0.0876 A cm-2). This work demonstrates a quantitative framework for corrosion-aware electrolyte optimization relevant to aluminum-based electrochemical systems.