AI-Based Cryptography: A Comprehensive Review of Adaptive Encryption, Neural Cryptanalysis, and Post-Quantum Resilience

Abstract

Nowadays, protecting information is paramount. With the rapid evolution of cyber threats and the growing potential of quantum computing to compromise classical encryption methods, the integration of Artificial Intelligence (AI) into cryptography has become a promising area for exploration. This review outlines recent advances in AI-based encryption techniques, such as dynamic key generation, adversarial neural cryptography, and AI-enhanced cryptanalysis. It compares AI-based approaches to traditional cryptographic systems in terms of adaptability, security strength, and quantum resistance. This paper also summarizes key research contributions and highlights the benefits and current challenges of deploying AI-based cryptography, including computational overhead and the absence of formal security guarantees. The review aims to offer an introductory overview of how AI is reshaping the future of encryption and to identify areas that require further research and validation.