Background <p>Artificial intelligence (AI) is increasingly being implemented in digital pathology to support the tissue classification, cell detection, biomarker quantification, the grading and prediction of clinically relevant molecular alterations.</p> Objective <p>Current applications of AI in the pathology of internal and oncological diseases are summarized with a focus on the most important algorithm approaches, representative cases of diagnostic applications and current limits of clinical implementation.</p> Material and methods <p>This narrative overview of the most recent advances describes the essential model forms, including tissue segmentation systems, algorithms for recognition of individual cells, unsupervised learning and foundation models, tools for the evaluation of immunohistochemistry and multimodal or language-based applications. Representative studies on renal, liver, pulmonary, gastrointestinal, hematological and thyroid gland pathologies are discussed.</p> Results <p>In many situations AI improves the reproducibility, objectiveness and efficiency. Some systems achieve an accuracy that is comparable to that of experts. Most AI tools are still in the validation stage and only a few have been transferred to routine clinical use.</p> Discussion <p>Artificial intelligence could further optimize the diagnostic and predictive histopathology. Their role remains assistive. A broad implementation is limited due to various hurdles and bottlenecks. The future progress depends on prospective validation and the integration into routine digital workflows.</p>