This MS is dedicated to numerical methods for multiscale modelling of processes, materials and structures with AI, model order reduction techniques, or data-driven approaches. Material modelling can induce high complexity and/or intractable computational times, especially for difficult problems such as multiscale modelling with nonlinear behaviours, fine resolutions, real-time simulations of processes and structures with complex behaviours, among others. The objective of this MS is to bring together researchers and specialists for discussing the most recent advances in this area. Topics include, but are not limited to: the construction of surrogate models for material and complex systems prediction, data-driven prediction of mechanical behaviours, material design and topology optimization accelerated by AI/model reduction, multiscale optimization and digital twins of processes, or construction of digital twins coupled with complex material modelling.