Developing and Designing a Secure Framework for Virtual ATM in Modern Banking Systems Using Blockchain Technology and Artificial Intelligence

The design, implementation and empirical assessment of a secure framework for a Virtual ATM using Hyperledger Fabric blockchain technology integrated with AI based multimodal biometric authenticator have been presented in this paper. The empirical assessment of 10,000 simulated banking sessions, indicated unequivocally this framework met the stringent financial authentication requirements for all of the evaluated dimensions. The quantitative equivalent of this study's key findings are: (1) Multi-modal fusion authentication provided an accuracy of 98.7%, with Precision, Recall and F1-Score each at 98.7%—all of which surpassed all individual biometric modalities (Face-95.6%, Voice-96.3%, Behavioural-94.2%); (2) The AUC-ROC for the Fusion System was 0.997 and the Average Precision was 0.989, compared to AUC values of 0.972-0.988 for each of the individual biometric modalities; (3) The Fusion System FAR was 0.0026%, which was a 769 fold improvement over traditional card and PIN authentication systems and FRR was 0.3% at the banking recommended operating point (time to authorise transaction), which represented a 73% reduction in FRR compared to working with only Face Recognition technology; (4) The Liveness Detection System was 99.4% accurate in detecting antagonistic spoofs for 847, using photograph (99.6%), video replay (99.2%) and 3D mask (99.3%) anti-spoofing testing; as such, this system satisfied the ISO/IEC 30107-3 Level 2 PAD requirements; (5) End-to-end transaction latency was 1.8 seconds under baseline operational conditions and 2.9 seconds for a sample size of 500 concurrent users, satisfying the second threshold for acceptable customer facing financial applications; and (6) No high risk penetration testing vulnerabilities were detected during the 847 vector penetration testing and 100% assurance that all Blockchains used in this study were unaltered; i.e., were immune to tampering. The convergence of blockchain and AI in the V-ATM framework represents a fundamental architectural paradigm shift from hardware-dependent, centralized infrastructure to software-defined, decentralized, intelligence-augmented financial services. As financial inclusion imperatives drive banking services into underserved populations globally, eliminating capital-intensive ATM hardware while enhancing security properties offers compelling economic and social value. Future research directions include: (1) implementation of post-quantum cryptographic primitives (CRYSTALS-Dilithium) for long-term quantum threat resistance; (2) offline transaction capability using cryptographic voucher mechanisms; (3) cross-jurisdictional regulatory compliance analysis for multi-national deployment.

Abubakar, M., Jaroucheh, Z., Al-Hashimi, M., & Liu, W. (2019). Decentralized trustless IoT data integrity verification using Ethereum blockchain. In Proceedings of the 2nd International Conference on Blockchain Technology (pp. 62–67) IEEE. https://doi.org/10.1109/BlockTech47271.2019 Ali, O., Ally, M., Clutterbuck, P., & Dwivedi, Y. K. (2020). The state of play of blockchain technology in the financial services sector: A systematic literature review. International Journal of Information Management, 54, 102199. https://doi.org/10.1016/j.ijinfomgt.2020.102199 Bhattacharyya, S., Jha, S., Tharakunnel, K., & Westland, J. C. (2019). Data mining for credit card fraud: A comparative study. Decision Support Systems, 50(3), 602–613. https://doi.org/10.1016/j.dss.2010.08.008 Chaum, D. (1983). Blind signatures for untraceable payments. In D. Chaum, R. L. Rivest, & A. T. Sherman (Eds.), Advances in Cryptology: Proceedings of Crypto '82 (pp. 199–203). Springer. https://doi.org/10.1007/978-1-4757-0602-4_18 Chen, L., Lee, W.-K., Chang, C.-C., Choo, K.-K. R., & Zhang, N. (2022). Blockchain based searchable encryption for electronic health record sharing. Future Generation Computer Systems, 95, 420– 429. https://doi.org/10.1016/j.future.2019.01.024 European Association for Secure Transactions (EAST). (2023). European ATM crime report 2023. EAST. https://www.association-secure-transactions.eu/industry-information/atm-crime-reports Ibrahim, M., & Al-Turjman, F. (2024). 5G-enabled smart banking kiosks: Edge-assisted biometric