Systematic Literature Review: Risk Assessment of AES for Secure Data Communication: Capabilities, Limitations, and Research Directions

Abstract

The most widely used symmetric encryption technique for protecting digital communications in a variety of applications, including cloud infrastructure and Internet of Things (IoT) devices, is still the Advanced Encryption Standard (AES). Even though AES has been widely used for 20 years and has a strong mathematical design (Daemen & Rijmen, 1998; NIST, 2023), side-channel exploitation, developments in quantum computing, and poor key management pose new risks to AES implementations (Altigani et al., 2021; Dheeba et al., 2025). The five crucial elements of AES security concerns are evaluated in this systematic literature review: implementation vulnerabilities, side-channel attack susceptibility, quantum computing dangers, key management issues, and AI-based attack detection. We examined 95 peer-reviewed papers from 2021 to 2025 in accordance with accepted systematic literature review guidelines for computer science research (Kitchenham et al., 2009; Kitchenham & Charters, 2007). According to the assessment, implementation issues lead to serious real-world vulnerabilities even though AES maintains strong mathematical security against conventional cryptanalysis (Bogdanov et al., 2011). With more than 50,000 traces, side-channel attacks especially those enabled by deep learning can effectively get keys from disguised implementations (Kuroda et al., 2021). Because Grover's approach essentially reduces AES-128 to 64-bit security (Grassl et al., 2016), quantum computing presents unequal risks. The most important operational vulnerability is key management, since attacks that go around AES's mathematical strength are made possible by lifecycle errors (Shakor et al., 2024). Although AI-based detection has potential for spotting cryptographic abuse, it has trouble generalizing across many implementations (Ghimire et al., 2024). In addition to identifying priority research directions for post-quantum transition, hardware-software co-design, and automated key management, this analysis synthesizes findings to offer evidence-based recommendations for safeguarding AES deployments.