STAT3 is a transcription factor that regulates cell proliferation, metastasis, angiogenesis, and immune response. It is activated primarily by phosphorylation, which transmits signals from cytokines and growth factor receptors to the nucleus. STAT3 hyperactivation is common in human cancers and associated with a poor prognosis. The STAT3 signaling pathway is an important target for cancer therapies because of its role in tumor growth, metastasis, and drug resistance. The Src Homology Domain 2 (SH2) is required for STAT3 activation, receptor recruitment, and dimer formation, making it an important therapeutic target. The purpose of this study was to predict the effects of mutations on protein stability and flexibility in silico and to identify novel inhibitors targeting STAT3’s SH2 domain using virtual screening and docking. Twenty mutations were analyzed, with four (E616V, K591N, R609T, and S611I) chosen for further investigation. Binding DB yielded 1283 potential STAT3 inhibitors, eight of which were selected based on toxicity, Lipinski and Veber criteria, and IC50 values. Docking results revealed two compounds (49838879 and 135241747) with strong SH2 interactions and promising pharmacological profiles, indicating potential targeted therapies. The use of artificial intelligence, such as deep learning, could transform the field by addressing challenges such as accurately predicting mutation effects and identifying high-affinity inhibitors, thereby advancing therapeutic development.