Artificial intelligence for cryptography: a systematic literature review of methods, applications, and challenges

Abstract

Purpose: to identify, classify, and synthesize the application of artificial intelligence (AI) techniques in cryptography, including AI methods, cryptographic problems and applications, evaluation approaches, challenges, and emerging future research directions.

Method: this study employed a Systematic Literature Review (SLR) following the PRISMA framework. Studies were retrieved from Scopus and analyzed through screening, quality assessment, data extraction, and thematic synthesis based on five research questions.

Findings: neural networks, deep learning, and machine learning are the dominant AI techniques applied to cryptography, particularly in image encryption, privacy-preserving computation, healthcare, and IoT. Key challenges include computational overhead, methodological heterogeneity, reproducibility, and security validation, while emerging directions emphasize adaptive, privacy-preserving, and intelligent cryptographic systems.

Implications: provides an integrated framework for understanding AI applications in cryptography and identifies key methodological patterns, challenges, and research opportunities to guide the development of secure, efficient, and adaptive cryptographic systems.

Originality: lies in providing an integrated synthesis that systematically connects AI techniques, cryptographic applications, evaluation approaches, challenges, and emerging research directions within a unified framework.