The field of AI research is witnessing a significant shift towards multimodal learning, where diverse data sources are integrated to improve model performance. This approach has shown great promise in various applications, including agriculture, remote sensing, and animal welfare assessment. A key challenge in multimodal learning is ensuring model interpretability, which is crucial for understanding complex relationships between different data modalities. Recent studies have explored the use of attention mechanisms, feature attribution methods, and explainable AI frameworks to provide insights into model behavior. Notably, the integration of attention weights into explanation frameworks has been shown to enhance model explainability. Furthermore, multimodal AI systems have been successfully applied to real-world problems, such as crop yield prediction, animal health monitoring, and lameness detection in dairy cows. Overall, the field is moving towards the development of more transparent, explainable, and robust AI systems that can effectively leverage diverse data sources. Noteworthy papers include: Intrinsic Explainability of Multimodal Learning for Crop Yield Prediction, which leverages Transformer-based models to explain multimodal learning networks. Can Multitask Learning Enhance Model Explainability, which explores the use of multitask learning to intrinsically explain model behavior. Multimodal AI Systems for Enhanced Laying Hen Welfare Assessment and Productivity Optimization, which highlights the potential of multimodal AI in animal welfare assessment.