Passive acoustic monitoring is a well-established, non-invasive technique for wildlife monitoring, with growing interest in bioacoustic individual-level recognition—the ability to distinguish individual animals based on their vocalisations. While existing approaches perform well when all individuals are known a priori, wildlife populations are inherently dynamic, making such closed-set assumptions unrealistic in natural environments. Recent work has begun to address this limitation by enabling models trained on a subset of known individuals to detect and recognise previously unseen individuals, a challenge known as out-of-distribution classification. Using novel datasets, these studies demonstrate that feature extractors pretrained for species classification can be successfully adapted to extend individual-level recognition beyond the training set, representing a crucial step towards real-world applicability. However, the decision-making processes of these models remain largely opaque, offering limited insight into why specific individuals are recognised or misclassified.