Safety and comfort of crew and personnel in harsh sea conditions are critical aspects of ship design. However, a comprehensive analysis of ship motions or at least a verification of vulnerability criteria from the second-generation intact stability criteria during the initial design stage are too demanding and time-consuming tasks. To address this challenge, we developed a surrogate model utilizing neural networks to streamline that assessment process. The model integrates a variety of parametric hull variants generated using a parametric CAD modeler, extensive acceleration criteria evaluations according to the second-generation intact stability criteria, and artificial neural networks. By training the neural network on data obtained from stability software, the surrogate model can rapidly predict stability characteristics. This approach significantly reduces the time required for exploration of design space and design assessments, enhancing efficiency and contributing to the design of safer, more comfortable vessels for harsh operating conditions.