Skin cancer is among the most common and potentially fatal forms of cancer worldwide, making early and accurate diagnosis essential for effective treatment and improved patient outcomes. Conventional diagnostic procedures, such as histopathological examination and biopsy, are often time-consuming and require significant clinical expertise. Recent advances in artificial intelligence and deep learning have enabled the development of automated diagnostic systems that can support clinicians in the early detection of skin lesions. In this study, a deep Convolutional Neural Network (CNN) is proposed for the classification of skin cancer images. To enhance the performance of the network, the Ant Colony Optimization (ACO) algorithm is employed to optimize the model parameters and reduce classification error. The proposed hybrid CNN–ACO framework is evaluated using a benchmark skin lesion dataset and compared with conventional deep learning approaches. Experimental results demonstrate that the optimization process improves classification performance and model convergence. The proposed method achieves an overall classification accuracy of 96%, outperforming baseline architectures and highlighting the effectiveness of integrating metaheuristic optimization techniques with deep learning models for automated skin cancer diagnosis. The findings suggest that the proposed framework can serve as a reliable decision-support tool for early skin cancer detection in clinical applications.