Extreme and rare events play an important role in multiple areas of engineering and science, including reliability and risk analysis, climate dynamics, disaster prediction and prevention, materials science, and countless others. Over the last several decades, there have been important developments in the modeling and understanding of such events and processes—from the mathematical, statistical, and algorithmic perspectives, but also in applications. New ideas and methods from AI have also brought fresh perspectives to the field, as have other data-driven methods, especially those developed for the low-data regime. Yet there remain important open questions and opportunities. For instance, the interaction between sampling and flow-based generative modeling has created powerful new approaches to simulation, but the performance and tailoring of these approaches to tails and extremes remains largely unexplored and without sufficient theoretical understanding. Another area of interest is the combination of generative modeling with statistical regularization algorithms that can generate unprecedented extreme events, i.e., events not contained in the training dataset, simply by relying on non-extreme event data and some statistics of observables. Physics-informed machine learning ideas have also seen widespread adoption in myriad fields, but a rigorous and quantitative understanding of their benefit in extreme event simulation is missing. These questions are made much richer when considering that model misspecification is inevitable in complex applications and especially in the low-data regime inherent to the rare event setting; hence, the prediction and characterization of tail events need sharper analyses of robustness and new methods that can guarantee robust performance or calibrated uncertainty quantification under realistic assumptions. Game-theoretic approaches and optimal transport tools have been adopted in recent years, but algorithmic and modeling issues involved in using such tools to inform and quantify model misspecification in rare event settings remain largely unexplored. Talks and discussions in this workshop will explore all of these open issues and the future of the field.