International Journal of Multidisciplinary Research and Growth Evaluation www.allmultidisciplinaryjournal.

com

AI-Driven cyber security: Security intelligence modelling

Guru Prasad BS 1*, Dr. Kiran GM 2, Dr. Dinesha HA 3
1
Student, Shridevi Institute of Engineering and Technology, Tumakuru, Karnataka, India
2
Associate Professor, Department of CSE, Shridevi Institute of Engineering and Technology ,Tumakuru, Karnataka, India
3
Professor, Dept. of CSE, Shridevi Institute of Engineering and Technology, Tumakuru, Karnataka, India

* Corresponding Author: Guru Prasad BS

Article Info Abstract

The process of defending computer networks from cyber attacks or unintended,
unauthorized access is known as cyber security. Organizations, businesses, and
ISSN (online): 2582-7138 governments need cyber security solutions because cyber criminals pose a threat to
Impact Factor: 5.307 (SJIF) everyone. Artificial intelligence promises to be a great solution for this. Security
Volume: 04 experts are better able to defend vulnerable networks and data from cyber attackers by
Issue: 06 combining the strengths of artificial intelligence and cyber security. This paper
November-December 2023 provides an introduction to the use of artificial intelligence in cyber security. AI-driven
cyber security refers to the use of artificial intelligence and machine learning
Received: 05-10-2023; technologies to enhance the protection of computer systems and networks from cyber
Accepted: 07-11-2023 threats such as hacking, malware, phishing, and other forms of cyberattacks. AI-
Page No: 961-965 powered security solutions are designed to automate the process of detecting,
analyzing, and responding to security incidents in real-time, thereby improving the
efficiency and effectiveness of cyber defense. These solutions can analyze large
amounts of data, identify patterns and anomalies, and make decisions faster and more
accurately than humans alone, enabling organizations to stay ahead of evolving cyber
threats.

DOI: https://doi.org/10.54660/.IJMRGE.2023.4.6.961-965

Keywords: Cyber security, Cyber-attacks, Artificial Intelligence

Introduction
Cybercriminals pose a significant threat to the online world, and it is essential to take necessary precautions to protect se nsitive
information. It is crucial for organizations, businesses, governments, and individuals to understand the risks associated w ith online
activities and to implement measures to reduce these risks. This includes using strong passwords, regularly updating software and
hardware, keeping software up-to-date, and backing up important data. Additionally, it is recommended to educate employees and
family members about safe online practices and to be aware of common scams and phishing attempts. By being proactive in prote cting
personal and sensitive information, the risk of cyber-attacks can be minimized and individuals and organizations can secure their online
presence. Cyber security refers to technology and practices designed to protect networks and information from damage or unauthorized
access [5]. It is vital because governments, companies, and military organizations collect, process, and store a lot of data. Cyber security
takes different forms, including military, law enforcement, judicial, commerce, infrastructure, and interior, intelligence, a nd
information systems. Cyber security is a dynamic, interdisciplinary field involving information systems, computer science, and
criminology. The security objectives have been availability, authentication, confidentiality, no-repudiation, and integrity [7].
The management of organizations should adopt a holistic approach to cyber security that covers people, processes, and technology.
This includes regular security training for employees, risk assessments, and incident response planning. The use of encryptio n and
multi-factor authentication can also help to secure sensitive information. In conclusion, managing cyber risks requires a comprehensive
and proactive approach that involves the entire organization [8].
By adopting best practices and staying informed about the latest threats and trends, organizations can reduce the risk of cyber-
attacks and ensure the safety of their data and systems.

961 | P a g e
International Journal of Multidisciplinary Research and Growth Evaluation www.allmultidisciplinaryjournal.com

Challenges the good outweighs the bad. Just as AI technologies can be

Although artificial intelligence tools could help fight used to identify and stop cyber-attacks, cybercriminals can
cybercrime, the tools are not a silver bullet and could be also use the AI systems to launch attacks. Besides, a shortage
exploited by malicious hackers. There are limitations that of cyber security experts is another problem. These
prevent AI from becoming a mainstream tool. The downsides challenges prevent AI from becoming the only cyber security
of AI in cyber security include cost, intensive resources, and solution.
training. AI in cyber security necessitates more resources and
funds than traditional, non-AI cyber security solutions, and it AI for Cyber Security
may be impractical in some cases. Cyber security is a domain The study concludes by emphasizing the importance of XAI
where absolute security is impossible. If a machine learning- in improving the transparency and accountability of AI
based security tool misses a particular kind of cyber-attack models in the field of cyber security, thereby enabling
because it is not coded into it that may lead to problems [4]. organizations to better understand and manage the risks
Hackers themselves can use AI to test and develop their associated with AI-based cyber security solutions [2]. The
malware and make it potentially AI-proof. Some critics have authors hope that the review will provide a valuable resource
warned that AI could make cyber-attacks more dangerous for researchers, practitioners, and stakeholders in the field of
and difficult to spot than ever before [1]. Some regard AI in XAI and cyber security, and will encourage further research
cyber security as posing both a blessing and a curse, although and development in this area [3].

Fig 1: Relationship between AI and ML

AI limitations necessitate the use of XAI in cyber security evades the AI-based detection framework. This highlights the
AI in cyber security has its own set of challenges and need for continuous improvement and adaptation of AI
limitations. Evasion attacks are one of the major concerns models to stay ahead of attackers and provide effective cyber
where attackers can manipulate the malware in a way that it security solutions [10].

Fig 2: AI cyber incidents detection and response

962 | P a g e
International Journal of Multidisciplinary Research and Growth Evaluation www.allmultidisciplinaryjournal.com

AI-driven cyber security refers to the use of artificial responding to security incidents in real-time, thereby
intelligence and machine learning technologies to enhance improving the efficiency and effectiveness of cyber defense.
the protection of computer systems and networks from cyber These solutions can analyze large amounts of data, identify
threats such as hacking, malware, phishing, and other forms patterns and anomalies, and make decisions faster and more
of cyberattacks. AI-powered security solutions are designed accurately than humans alone, enabling organizations to stay
to automate the process of detecting, analyzing, and ahead of evolving cyber threats [3].

Fig 3: Schematic representation of the AI architecture and its attack surface

AI Cyber security Solutions stories, AI technologies like machine learning and natural
As cyberattacks grow in volume and complexity, artificial language processing provide rapid insights to cut through the
intelligence (AI) is helping under-resourced security noise of daily alerts, drastically reducing response times.
operations analysts stay ahead of threats. Curating threat Watch the video to see how AI helps analysts connect the dots
intelligence from millions of research papers, blogs and news between threats [6].

Fig 4: Application of AI across the cyber kill chain

Enabling AI-powered Smarter Cyber security Solutions Security Intelligence, Operations and Analytics, Malware
Service offerings for Enabling AI-driven Cyber Security for Detection and Vulnerability management Solutions.

Table 1: Results of website legitimacy decision – Database 1

Database 1
Decision
Phishing Non-phishing Suspicious
Phishing 567 256 212
Non-Phishing 284 159 369
Suspicious 216 241 152

963 | P a g e
International Journal of Multidisciplinary Research and Growth Evaluation www.allmultidisciplinaryjournal.com

Fig 5

Table 2: Results of website legitimacy decision – Database 2

Database 2
Decision
Phishing Non-phishing Suspicious
Phishing 201 135 189
Non-Phishing 125 125 152
Suspicious 157 141 128

Fig 6

Conclusion machine learning, 2019, 1210-1214.

AI can detect and stop cyber threats in real-time with limited 3. Baptista I, Shiaeles S, Kolokotronis N. A novel malware
resources. The constantly evolving nature of cyber-attacks detection system based on machine learning and binary
means that humans shall struggle to keep up with the intel. visualization, 2019, 1-6.
However, using machine learning, AI can chomp down data https://doi.org/10.1109/ICCW.2019.8757060
for quick analysis and provide excellent security coverage 4. Barbara D, Couto J, Jajodia S, Popyack L, Wu N. Adam:
without taking much time or energy away from the existing Detecting intru-sions by data mining, 2001, 5-6.
tasks. Machine learning allows Human analysts to focus on 5. Bose S, Barao T, Liu X. Explaining ai for malware
interpreting the results from deep analysis and devising novel detection: Analysis of mechanisms of malconv. In: 2020
techniques for fighting cyber-crime. International Joint Conference on Neural Networks
(IJCNN), 2020, 1-8.
References 6. Chowdhury M, Rahman A, Islam MR. Malware analysis
1. Al-Yaseen W, Othman Z, Ahmad Nazri MZ. Multi-level and detection using data mining and machine learning
hybrid support vector machine and extreme learning classification, 2018, 266-274.
machine based on modified k-means for intrusion https://doi.org/10.1007/978-3-319-67071-3 33
detection system. Expert Systems with Applications 67 7. Coull S, Gardner C. Activation analysis of a byte-based
(01 2017) deep neural network for malware classification, 2019,
2. Banu R, MA, CA, SA, Ujwala H, NH. Detecting 21-27. https://doi.org/10.1109/SPW.2019.00017
phishing attacks using natural language processing and 8. Demetrio L, Biggio B, Lagorio G, Roli F, Armando A.

964 | P a g e
International Journal of Multidisciplinary Research and Growth Evaluation www.allmultidisciplinaryjournal.com

Explaining vulnera-bilities of deep learning to

adversarial malware binaries, 2019.
9. Feng F, Zhou Q, Shen Z, Xuhui Y, Lihong H, Wang J.
The application of a novel neural network in the
detection of phishing websites. Journal of Ambient
Intelligence and Humanized Computing, 2018.
10. Feng W, Sun J, Zhang L, Cao C, Yang Q. A support
vector machine based naive bayes algorithm for spam
filtering, 2016, 1-8.

965 | P a g e