Designing composite structures for precise applications is challenging due to their anisotropic and inhomogeneous properties, requiring cost-effective and time-efficient analysis methods to ensure robustness. To address this, a predictive model based on Artificial Neural Network (ANN) was developed to predict the buckling load of hybrid composite laminate designs under uniaxial compression. Seventeen sample designs were generated using a Design of Experiment (DOE) method incorporating combinations of key parameters (angle orientation, volume fraction, and plate thickness) and tested using finite element simulations in ANSYS. The Box-Behnken Design (BBD) method was used for DOE to optimise the sample configurations. Graphite/Glass epoxy hybrid composites analysed in an eight-layer symmetrical lamination pattern [0/+θ/−θ/90]s, with orientation angles of 0° to 45°, plate thicknesses of 0.5 mm, 2 mm, and 3 mm, and volume fractions of 100% graphite, 100% glass, and a hybrid mix. The optimal design was identified as a full graphite laminate with a 22.5° fibre orientation and 3 mm thickness, achieving the highest buckling load capacity. The ANN model, trained with Levenberg-Marquardt algorithm achieved high prediction accuracy with an R2 of 0.93976. This work enhanced the understanding of buckling behaviour in hybrid composite laminates under uniaxial compression and evaluates the accuracy of ANN predictions against finite element analysis results.