In this research, an artificial neural network model was developed to predict the credit risk of clients in the financial system, surpassing traditional methods in accuracy, sensitivity, and specificity. Using data from Caja Sullana, which included 85 client records with variables such as income, credit history, and age, a cross-sectional non-experimental design was employed. The network, composed of an input layer, a hidden layer, and an output layer, was trained using a scaled conjugate gradient algorithm, standing out for its efficient and simple architecture. The results showed a 100% accuracy in training and testing, and a general accuracy of 98.25%, with outstanding sensitivity and specificity, thus validating the proposed hypotheses. This study demonstrates the capability of neural networks to significantly enhance credit risk evaluation, offering a powerful tool for credit decisi...