New and emerging technologies have always been used in climate science, and current trends in advanced computing and new data and data driven methods are no exception. Advanced modelling efforts seeking to represent the global Earth system in ever finer detail are targeting cloud-resolving, km-scale resolutions using the latest computing architectures. At the same time, there have been advances in novel observations relevant to such high resolution model processes and model observation simulators and in new classes of observations from distributed sensors or satellite constellations. Machine Learning and Artificial Intelligence approaches are now being integrated into Earth System Models (ESM) and earth observation frameworks, and being used increasingly to augment or emulate models in various ways. Increased resolution and more complex process representation in ESMs has implications for observations that are required to initialise, evaluate, and develop traditional ESMs. New data requirements for training, validating, and critically assessing biases in AI models and model emulators are increasingly incorporating a diverse range of Earth Observation (EO) data. Similarly, ESM use cases are driving development for upcoming or proposed EO missions as well as enabling reprocessing of preexisting data for process understanding. Such advancements have highlighted the need for improvements in communication between these two pillars of climate research and identified new avenues for collaboration. In this session organised by WCRP ESMO and Digital Earth LHA, we invite presentations that discuss the fusion of models and observations, especially those using new technologies such as AI and exascale computing. The presentations will include contributions from several WCRP projects and will help guide a later discussion on improving collaborations in the topic.