Zero-Trust Access Control Via Blockchain-Backed Identity and AI-Based Anomaly Detection

Security models need to expand beyond conventional perimeter-based defenses to protect modern organizational networks from progressively sophisticated insider threats. However, current systems are unable to address the crucial gaps in Zero-Trust implementation because they are either unable to confirm identity integrity or identify behavioral abnormalities in real time. We propose BlockZTA-AI, a blockchain-backed, AI-driven Zero-Trust access control framework designed to enforce adaptive anomaly detection and continuous, tamper-proof authentication, as a solution to this problem. The model combines an LSTM-Autoencoder anomaly engine, federated threatintelligence learning, contextual risk scoring, and a cryptographic identity ledger. BlockZTA-AI outperforms baselines like Random Forest, Gradient Boosting, LSTM, and GRU by achieving 98.97% accuracy, 98% precision, 99% recall, and 98% F1-score using the CERT Insider Threat dataset, which contains 2.63 million emails and comprehensive psychometric features. Our main contributions are threefold: a completely decentralized identity validation mechanism, a privacy-preserving federated learning protocol enhanced with blockchain trust scores, and a real-time adaptive anomaly detection pipeline that strengthens Zero-Trust enforcement in dynamic organizational settings.