CC-FEAI: A Compliance-Centric Federated and Explainable AI Architecture for Regulated Domains

This journal-format author manuscript presents CC-FEAI, a compliance-centric architecture integrating federated learning, explainable AI, privacy protection, security controls, regulatory evidence and lifecycle governance for regulated domains. Derived from the author’s doctoral thesis, the study uses a design-science approach and controlled evaluation to examine predictive performance, explanation artefacts and compliance-readiness controls. The manuscript reports FedAvgM results of 95% accuracy, 94% precision, 92% recall and 93% F1 in the tested configuration. These results demonstrate controlled technical feasibility and do not constitute journal peer review, legal certification or production validation. Authored and authorised by Dr. Jami Kiran Kumar.