Highly pathogenic avian influenza (HPAI) represents a threat to animal health, human health, and economic prosperity, with the ongoing outbreak in wild and domestic animals since 2021 being the largest on record. This outbreak is associated with the 2.3.4.4b clade of influenza A, and it is as yet unclear what factors have contributed to its spread at the continental scale. In this study we use Bayesian additive regression trees, a machine learning method designed for probabilistic modelling of complex nonlinear phenomena, to construct species distribution models for HPAI presence across Europe. Using these models we identify factors driving the geospatial distribution of cases and project the distribution of risk across Europe. Our models are stratified by time to capture both seasonal changes in risk patterns and shifts in HPAI epidemiology associated with the introduction of the 2.3.4.4b clade. While previous studies have aimed to predict HPAI presence from physical geography, here we explicitly consider the impact of wild bird ecology by including in our model estimates of bird species richness, abundance of specific high-risk bird taxa, and “species-trait abundance indices” describing the total abundance of species with high-risk behavioural and/or dietary traits. Our projections point to a shift in concentration of risk towards cold, low-lying regions of coastal northwest Europe associated with 2.3.4.4b, with the margins of uncertainty extending that risk further into central and eastern Europe. In coastal northwest Europe specifically, we predict a persistence of high risk throughout the year. Methodologically, we demonstrate that while the majority of variation in risk can be explained by climate and other aspects of physical geography, the addition of ecological covariates represents a valuable refinement to species distribution models of HPAI.