Equation discovery from data (EQD) or symbolic regression (SR) is a task for artificial intelligence (AI) methods rooted in machine learning and symbolic computation and has garnered significant attention within the realm of computational mechanics and material science. It enables data-driven modelling that is inherently interpretable, i.e., producing equations that best describe a dataset. Because computational mechanics is built around mathematical models, the equations produced by these AI methods are readily incorporated into a variety of preexisting workflows, including analytical derivations used in computational applications. This minisymposium seeks to illuminate the latest breakthroughs and applications of EQD and SR techniques in advancing the simulation, analysis, and optimization of complex mechanical systems.      By harnessing the potential of AI for equation discovery, researchers are discovering novel avenues to enhance accuracy, model interpretability, and computational efficiency in the domain of computational mechanics. This minisymposium will explore a comprehensive array of topics that include, but are not limited to: • Equation Discovery and System Identification: AI techniques and SR enable the discovery of governing equations from experimental or simulated data, facilitating the identification of system dynamics and behaviors without a priori assumptions. • Physics-Informed Machine Learning: Integrating domain-specific physical insights into EQD methods may yield hybrid models that combine data-driven learning with established physical laws to enable identification of the physical significance of individual model components while also ensuring greater interpretability, model generalization and robustness. • Uncertainty Quantification, Uncertainty Propagation and Sensitivity Analysis: Techniques for uncertainty quantification and propagation of uncertainties and sensitivities for equation discovery methods and SR within computational mechanics models, providing a deeper understanding of system behavior through traceable uncertainty propagation. • Equation Discovery and SR in Multiscale Modelling: Computational mechanics and material sciences require modelling effects across multiple scales. EQD and SR are capable of producing efficient and accurate surrogates or emulators for computationally expensive numerical models. • Optimization, High-throughput Screening, and Design: The application of EQD and SR techniques in optimization sce