We investigated factors affecting avian influenza virus (AIV) detections in Danish wild birds using data from the passive and active AIV surveillance in wild birds from 2006-2020. We used this data and machine learning (ML) algorithms along with landscape and environmental variables to develop predictive models of AIV occurrence in Denmark. We furthermore assessed potential accessibility bias in the passive AIV surveillance data submitted by the public. The passive AIV surveillance data was biased regarding accessibility to areas compared to random locations within Denmark. ML models differed in their predictive power and were used to predict the risk of AIV presence throughout Denmark. Our results suggest that landscape variables may affect AIV presence and enabled us to create risk maps of AIV occurrence in Danish wild birds. This may aid future targeted surveillance efforts within Denmark.