The fatigue behavior of metallic alloys, especially the 2024-T351 aluminum alloy, is a major concern in the aerospace industry, where component safety and durability are critical. Traditional fatigue life prediction methods rely on extensive and expensive experimental procedures. Moreover, the complex microstructural mechanisms involved make analytical modeling challenging. This study introduces a deep learning-based approach for predicting the fatigue life of 2024-T351 aluminum alloy under cyclic loading (R = 0,1). A Convolutional Neural Network (CNN) was trained to model crack propagation behavior using experimental data. The model achieved a high coefficient of determination (R2 = 0,978), demonstrating strong predictive capabilities even with a limited dataset. These results highlight the CNN's ability to capture complex fatigue-related patterns and generalize effectively. The proposed method offers a promising alternative to conventional testing by significantly reducing the need for physical experiments. It enables quicker and more cost-efficient assessments, with potential applications not only in aerospace but also in other industries where fatigue performance is critical. This work contributes to the growing integration of machine learning in materials science, offering a step forward in data-driven prediction of structural integrity.