Earthquake catalogues are a crucial resource for most seismological investigations and form the foundation of every probabilistic seismic hazard assessment (PSHA). This paper presents new developments in processing and analyzing an updated European earthquake catalogue for PSHA purposes. The updated catalogue expands on the work of the 2020 European Seismic Hazard Model (ESHM20) and now covers the period from 1000 to the end of 2021, contributing with an additional 4,500 revised earthquakes with magnitudes greater than 3.5 in the region. Using this dataset, we test various state-of-the-art procedures for analyzing spatial patterns, magnitude distributions, and seismicity rates. In particular, we explore the use of Machine Learning (ML) techniques to address long-standing problems, such as earthquake declustering. We introduce a new objective tectonic classification methodology that improves on current standards and enables the grouping of previously unclassified events via an iterative clustering approach. This work lays the foundation for creating a new standardized workflow for developing the next generation of the Moody’s RMS European seismic hazard model.