International Journal of Science and Research (IJSR)

International Journal of Science and Research (IJSR)
Call for Papers | Fully Refereed | Open Access | Double Blind Peer Reviewed

ISSN: 2319-7064




Downloads: 7 | Views: 272 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1

Research Paper | Electronics & Communication Engineering | India | Volume 12 Issue 8, August 2023 | Rating: 5 / 10


Autonomous Cyber Systems Using AI - Approach on How to Improve Detection and Response

Badri S.


Abstract: Industrialization of Cyber Attacks is leading to the increase in their severity and complexity. Organizations around the globe are moving to adopting AI to improve their InfoSec posture. However, despite the measures taken the protective mechanisms are not adequate. The industrialization has lead to the emergence of easily available hacking tools and services online as well. State sponsored, motivated activists and young enthusiasts are constantly using these tools & services. Data breaches and Ransomwares have grown exponentially. Ransoms paid do not necessarily lead to complete decryption. Payouts are demanded through stealth crypto currency like Monero. Innocent clicks by ignorant internal users in organizations cause major business breakdown no matter how complicated the cyber defenses that are implemented. There is initial work which has been done on Autonomous Cyber Defense (ACyD / ACD) by re-searchers to explore possible mechanisms of defense. However, there are certain limitations in the approaches. This research proposal is focused on strengthening the existing research that has been done, through some proven mechanism and techniques. These autonomous actions described in these researches uses AI, inputs for which are based on ML, DL, CNN, NLP and Visual AI. The outcomes however can be improved to protect data at rest, in motion and during consumption through techniques discussed in this research proposal. Data in the form of IP addresses, SIEM data, User & Entity Behavior Data, Existing Malwares etc. are necessary to evolve some models. Better Data classification, Clustering etc would help improve existing some models such that the accuracy to identify anomalies and taking necessary self-defending actions are better. Both InfoSec service providers / organizations, would then be enabled to defend & respond better, and this research proposal is focused on that.


Keywords: AI, EDR, Cyber Attack, InfoSec Posture, Cyber Defense, ML, UEBA, Vulnerability, ACyD, ACD, ACO, AI, DL, CNN, GAN, NLP


Edition: Volume 12 Issue 8, August 2023,


Pages: 426 - 433


How to Download this Article?

Type Your Valid Email Address below to Receive the Article PDF Link


Verification Code will appear in 2 Seconds ... Wait

Top