Deep learning is a state-of-the-art pattern recognition technique that has proven extremely powerful for the analysis of digitized histopathological slides. In our work, we explored the application of deep learning across various pathology diagnostics, aiming to support pathologists by increasing accuracy and consistency while reducing workload. This presentation will address the existing translation gap: despite the development of numerous AI models, few have reached clinical implementation. We propose that the next critical step involves large-scale validation with pathologists to evaluate the added value of these algorithms in routine practice. Such studies are essential for certification but are currently lacking in our field, hindering the assessment of deep learning's true potential in pathology diagnostics.