Abstract Artificial intelligence (AI) has become an increasingly important tool in dermatology, largely due to the visual nature of skin disease assessment and the rapid development of machine learning techniques. In particular, deep learning models based on convolutional neural networks have demonstrated high performance in the analysis of clinical and dermoscopic images, enabling automated classification, segmentation and detection of skin lesions. Numerous studies report diagnostic accuracy comparable to that of experienced dermatologists in controlled experimental settings. Recent research extends beyond retrospective validation and emphasizes clinical implementation, human–AI collaboration and integration into diagnostic workflows, including total body photography and longitudinal surveillance of high-risk patients. Despite these advances, significant challenges persist, such as limited model generalizability, underrepresentation of darker skin phototypes, variability in image acquisition conditions and the limited interpretability of deep learning systems. This review summarizes current applications of AI in dermatology, highlights key limitations and outlines future directions for the safe, transparent and equitable integration of AI into routine clinical practice.
Zieleń et al. (Tue,) studied this question.