Breast cancer is the most common cancer in women worldwide and metastatic dissemination is the principal factor related to death by this disease. Breast cancer stem cells (bCSC) are thought to be responsible for metastasis and chemoresistance. In this study, based on whole transcriptome analysis from putative bCSC and reverse engineering of transcription control networks, we identified two networks associated with this phenotype. One controlled by SNAI2, TWIST1, BNC2, PRRX1 and TBX5 drives a mesenchymal or CSC-like phenotype. The second network is controlled by the SCML4, ZNF831, SP140 and IKZF3 transcription factors which correspond to immune response modulators. Immune response network expression is correlated with pathological response to chemotherapy, and in the Basal subtype is related to better recurrence-free survival. In patient-derived xenografts, the expression of these networks in patient tumours is predictive of engraftment success. Our findings point out a potential molecular mechanism underlying the balance between immune surveillance and EMT activation in breast cancer. This molecular mechanism may be useful to the development of new target therapies.