Adaptive Cybersecurity using AI-Powered Software Agents
DOI:
https://doi.org/10.15662/IJARCST.2025.0804003Keywords:
Adaptive Cybersecurity, Artificial Intelligence, Machine Learning, Real-Time Protection, Cyber Threats, Zero-Day Attacks, Advanced Threats, Autonomous Security, Cybersecurity Systems, AI ProtectionAbstract
The conventional type of defense mechanisms, which were traditionally based on static defense mechanisms, do not protect as cyber threats continue to become more sophisticated. The current paper presents an innovative solution that adopts the concept of adaptive cybersecurity, utilizing AI-enabled software agents to detect, learn, and respond to emerging threats in real-time. These agents can constantly enhance their defenses by using machine learning algorithms and provide greater defenses against advanced persistent threats, zero-day attacks, and insider threats. This paper describes the design of the AI agents, their deployment, and their behaviour in controlled conditions, with a focus on the main performance indicators, including detection rate, response time, and adaptability to novel threats. The results of the study show that AI agents may drastically decrease mean time to detect (MTTD) and mean time to respond (MTTR) to threats, and also reduce false positives. This study demonstrates how AI-based adaptive systems can address the limitations of conventional cybersecurity, offering a framework for further development and practical implementation.
References
1. Ahmed, M. I., & Kannan, G. (2021). Secure End to End Communications and Data Analytics in IoT Integrated Application Using IBM Watson IoT Platform. Wireless Personal Communications. https://doi.org/10.1007/s11277-021-08439-7
2. Aslan, Ö., Aktuğ, S. S., Okay, M. O., Yilmaz, A. A., & Akin, E. (2023). A Comprehensive Review of Cyber Security Vulnerabilities, Threats, Attacks, and Solutions. Electronics, 12(6), 1–42. https://doi.org/10.3390/electronics12061333
3. Ferdous, J., Islam, R., Mahboubi, A., & Islam, Z. (2023). A State-of-the-Art Review of Malware Attack Trends and Defense Mechanism. IEEE Access, 11, 121118–121141. https://doi.org/10.1109/access.2023.3328351
4. Nalage, P. (2025). Agentic Digital Twins: Self-Evolving Models for Autonomous Systems. Well Testing Journal, 34(S3), 227-244.
5. Nalage, P., & Kusuluru, B. K. R. (2024). Agentic Scientific Twins for Hypothesis Generation and Experiment Simulation. CINEFORUM, 64(2), 265-290.
6. Nguyen, T. T., & Reddi, V. J. (2021). Deep Reinforcement Learning for Cyber Security. IEEE Transactions on Neural Networks and Learning Systems, 34(8), 1–17. https://doi.org/10.1109/tnnls.2021.3121870
7. Paidy, P. (2023). Leveraging AI in Threat Modeling for Enhanced Application Security. 4, 57–66. https://doi.org/10.63282/3050-9262.ijaidsml-v4i2p107
8. Qumer, S. M., & Ikrama, S. (2022). Poppy Gustafsson: redefining cybersecurity through AI. The Case for Women, 1–38. https://doi.org/10.1108/cfw.2022.000001
9. Radanliev, P., De Roure, D., Page, K., Van Kleek, M., Santos, O., Maddox, L., Burnap, P., Anthi, E., & Maple, C. (2020). Design of a dynamic and self-adapting system, supported with artificial intelligence, machine learning and real-time intelligence for predictive cyber risk analytics in extreme environments – cyber risk in the colonisation of Mars. Safety in Extreme Environments, 2(3), 219–230. https://doi.org/10.1007/s42797-021-00025-1
10. Ramamoorthi, V. (2021). AI-Driven Cloud Resource Optimization Framework for Real-Time Allocation. Journal of Advanced Computing Systems, 1(1), 8–15. https://doi.org/10.69987/
11. Shahrouz, M., Nazari, A., & Moradi, S. (2023). AI-Driven Cybersecurity: Legal and Ethical Considerations in Autonomous Systems Protecting Digital Networks. Legal Studies in Digital Age, 2(1), 1–12. https://jlsda.com/index.php/lsda/article/view/7
