A key focus of bioacoustics is the study of fauna through the vocalizations animals produce. Traditionally, sound characteristics have been analyzed to assess variability in these vocalizations. With the advent of machine learning, significant progress has been made in recognizing the presence of different animals in recordings. However, as these methods have become easier to deploy, concerns arise regarding the reliability of the results, particularly when targeting abundance estimation, which is crucial for biodiversity studies. To address this, hybrid monitoring techniques will be developed to provide ground-truth evidence. Additionally, recent advancements in explainable AI aim to bridge the gap between machine-generated insights and real-world parameters. The application of these methods is expected to yield substantial progress, paving the way for practical use in this field.