Lightweight Cryptography-Based Generative Explainable AI With Multi-Factor Authentication Methods for Safe Cyber Transactions in Oil Sector

Abstract

This chapter proposes a novel multi-factor authentication (MFA) with six schemes, namely password salting/hashing, non-interactive zero-knowledge (NIZK) proofs, GPS-based validation, time-based one-time passwords (TOTP), DNA cryptography, and lightweight SPECK ciphers. Taken together, these elements address the deficiencies in prior authentication and achieve a tradeoff between security and computational efficiency. The system is verified by theoretical and experimental methods. Furthermore, it is theoretically examined under the Real-or-Random model (RoR) with generative/explainable Artificial Intelligence (AI)-driven cybersecurity and provides strong security guarantees in terms of unpredictability (even if reduced in certain security parameters) and defends against replay, insider misuse, brute-force key search attacks, as well as spoofing ones. The solution is developed in Java, and the system is empirically evaluated by performing 100 runs to examine essential performance features: randomness, determinism, stability, and scalability.