Cancer-associated fibroblasts promote tumor progression through growth facilitation, invasion, and immune evasion. This study investigated the impact of activated cancer-associated fibroblasts (aCAFs) on survival outcomes, immune response, and molecular pathways in distal bile duct (DBD) cancer. We analyzed 469 patients (418 from our cohort and 51 from The Cancer Genome Atlas) with DBD adenocarcinoma. aCAFs were evaluated using hematoxylin and eosin staining. We developed a machine learning-based survival prediction model incorporating aCAFs and clinicopathologic parameters. Additionally, we performed differential gene expression analysis, Disease Ontology analysis, gene set enrichment analysis, and in vitro drug screening of aCAFs-related genes. The presence of aCAFs significantly correlated with poor survival, advanced T and N stages, infiltrative growth pattern, lymphatic/perineural/adjacent organ invasion, and decreased tumor-infiltrating lymphocytes. aCAFs-related genes were associated with immune system functions, G protein-coupled receptor signaling, and metabolic conditions (diabetes, obesity, and abnormal C-peptide levels). In machine learning-based survival models, aCAFs emerged as a strong discriminator for survival prediction. In vitro drug screening revealed that refametinib suppressed the growth of DBD carcinoma cells expressing high levels of fibroblast activation protein-α. In conclusion, integration of machine learning and systems biology analyses identifies aCAFs as potential biomarkers for risk stratification and therapeutic targeting in DBD cancer.