This session is a platform for Neural Earth System Modelling, where we gather contributions aiming for the integration of artificial intelligence (AI) into Earth and climate science. AI techniques like deep learning are advancing towards a critical skill threshold, allowing not only automated data analyses but physics-informed prediction and interpretation of geophysical and climate processes. As such, this session focusses on novel applications of AI techniques and on methodological achievements towards coupled hybrids of AI and Earth system modelling. Applications in this context include hind- and forecasting of geophysical processes, climate projections, Earth system data downscaling and gap filling, or other physics closure approaches, where traditional statistical or numerical approaches are inferior or even not existent. In particular, we invite contributions on explainability and interpretability of AI techniques in Earth system science.