Citation

Lor, Chyanne Wen Qian and Yusof, Ibrahim and Muhamad Amin, Anang Hudaya (2026) AI-Powered Dynamic Encryption and Decryption Defense Model. Journal of Informatics and Web Engineering, 5 (2). p. 197. ISSN 2821-370X

Text sucheng,+2180-Article+Text-final.pdf - Published Version Restricted to Repository staff only Download (898kB)

Abstract

Static cryptographic systems that rely on fixed algorithms and predetermined keys have proven increasingly ineffective in addressing today’s adaptive and AI-powered cyber threats. This paper proposes an Artificial Intelligence (AI)-enabled dynamic encryption and decryption defence model designed to enhance cybersecurity through real-time threat classification and context-aware cryptographic response. The framework combines the Suricata intrusion detection engine with a Random Forest model developed using the CIC-IDS2017 dataset, allowing it to identify and categorize network anomalies into unified groups, including Denial-of-Service (DoS), Distributed Denial of Service (DDoS), Brute Force, and PortScan. Once threats are identified, the system dynamically selects an appropriate encryption scheme, which is AES-128, AES-192, AES-256, or ChaCha20, based on the severity level of the threat. This proportional encryption logic is implemented through a weighted random function, ensuring both computational efficiency and data confidentiality. Logs are periodically encrypted using a scheduled batch system, and any decryption is restricted to time-limited, read-only access, backed by SHA-256 hash verification and secure key storage outside the logging directory. In a simulated environment, the framework demonstrated reliable classification performance with an overall accuracy of 79%, consistent encryption and decryption operations, and high forensic traceability through structured logging. Automation mechanisms, such as Windows Task Scheduler integration and failure recovery logic, ensured robustness against execution overlaps and latency spikes. The proposed architecture is modular, scalable, and designed for potential deployment in enterprise or cloud environments where automated, intelligent cryptographic control is essential. Overall, this work contributes a practical and intelligent solution for real-time, threat-responsive encryption in modern cybersecurity infrastructures.

Item Type:	Article
Uncontrolled Keywords:	AI-Based Threat Classification, Dynamic Encryption, Suricata, AES Encryption, Intrusion Detection System, Forensic Integrity
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD28-70 Management. Industrial Management > HD30.2 Electronic data processing. Information technology. Including artificial intelligence and knowledge management
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	09 Jul 2026 03:18
Last Modified:	09 Jul 2026 03:18
URII:	http://shdl.mmu.edu.my/id/eprint/16328

Downloads

Edit (login required)