In this study, the influence of different aging temperatures on the tensile ductility and Intergranular Corrosion (IGC) performance was investigated by using SEM, TEM, U2-net, DL-EPR, and EIS. The U2-net was employed for efficient and accurate identification of the Laves phase in the alloy 625. In parallel, an extensive analysis was conducted to assess the influence of Laves phase precipitation on the tensile ductility of the alloy. Additionally, an investigation into the precipitated phase was undertaken to explore the correlation between tensile ductility and IGC. The research findings indicate that with increasing aging temperature (from 873 K to 993 K), the elongation initially slightly increases, then sharply decreases, and finally slightly increases again. Simultaneously, the sensitivity to Intergranular Corrosion first increases and then slightly decreases. Additionally, the trained U2-Net achieved an average accuracy rate of 99.3% on the validation dataset, demonstrating that deep learning networks can effectively and rapidly identify Laves phases in SEM images. DL-EPR results reveal a certain correlation between high Ra values and low elongation.