CC BY
99
ПРЕДСТАВЛЕНИЕ НАУЧНОЙ РАБОТЫ
THE IMPACT OF ARTIFICIAL INTELLIGENCE ON CYBERSECURITY
ОРЧЕСТВА
Mammadov Turgay, Baku Engineering University, Baku, Azerbaijan
E-mail: tmemmedov9@std. beu. edu. az
Abstract. As the volume and complexity of cyberattacks, Artificial Intelligence (AI) is needed to overcome limitations of humans in cybersecurity. AI algorithms learn to how to react to various scenarios by using training data.
The study analyzes the advantages and limitations of Artificial Intelligence on cybersecurity. Beyond its benefits, AI has some drawbacks in cybersecurity. It can sometimes make data management more difficult and cause certain security issues.
Security professionals are better able to protect sensitive data and networks from cyberattacks by combining cybersecurity with AI.
This article provides a brief overview of AI implementations in cybersecurity and discuss ways to improve defense mechanism to cyber-attacks.
Key words: artificial Intelligence, cybersecurity, cybercrime, threats.
INTRODUCTION
As we know, Artificial Intelligence plays an important role in different fields of our life such as cybersecurity, education, chemistry, medicine and so on.
AI and cybersecurity have numerous interdisciplinary connections. Expert systems, neural networks, machine learning, fuzzy logic, data mining and other AI tools are increasingly being used to detect and prevent cybercrime. There are several benefits and limitations of applying Artificial Intelligence in cybersecurity.
Benefits of using AI in cybersecurity
Continuously learning: AI enhances its ability to understand cybersecurity risks and attacks by consuming billions of data items.
AI reasoning: Artificial Intelligence uses AI reasoning to find threats more quickly. AI evaluates relationships between risks like insiders, malicious files in seconds or minutes.
The time it takes security analysts to make important decisions and mitigate threats is decreased thanks to curated risk analysis that AI offers.
Machine Learning assists in the battle against scams: Google utilizes machine learning algorithms to avoid phishing frauds and spam emails. Google algorithms assists in recognizing malicious emails in order to inform and protect users.
Limitations of using AI in cybersecurity
On the other hand, there are some restrictions that prevent Artificial Intelligence from being a common security tool.
ВЕСТНИК НАУКИ И ТВОРЧЕСТВА
Resources - Companies should invest a significant amount of time and money to create and maintain AI system.
Data privacy - AI enabled third-party organizations to acquire and analyze more data on us, resulting in increasing privacy and security concerns.
Target for hacker - In order to make their malware resistant to AI based detection systems, attackers test and develop it. Hackers use existing AI tools to launch most powerful attacks and target conventional security systems as well as AI-enhanced ones.
Data sets - In order to detect and predict cyber-attacks and respond appropriately, massive amount of data must be collected and trained. It may not be an issue for large firms that deal with large amounts of data.
Main Challenges in Cybersecurity
Despite the advancements in cybersecurity, attacks are growing increasingly deadly. The following are the primary challenges in cybersecurity:
Hybrid and remote working
The incident tracking is more challenging due to distance. To efficiently monitor occurrences across geographies, cybercrime specialists should overcome difficulties in infrastructure.
Hackers frequently conceal and modify their IP addresses.
Hackers employ a variety of tools, including proxy servers, Virtual Private Networks (VPN) and others. These tools assist hackers in remaining anonyms and undiscovered.
Manual threat hunting
Threat hunting can dramatically lower the likelihood of an attack by revealing vulnerabilities, but using several tools makes the process very-time consuming. Evidence collection involves a lot of human work, and it needs to be verified using several third-party systems.
Reactive nature
Companies can only solve problems after they have occurred. Security specialists face a significant problem in predicting attacks they before arising.
AI cybersecurity solutions
Security endpoints: the advent of remote working and endpoint interconnection brings new cybersecurity challenges. Endpoint security detect suspicious behavior and quickly fix it. Combining supervised and unsupervised machine learning is proven to be quite successful at identifying suspicious behavior and restricting or preventing access. The following image shows how they're implementing machine learning algorithms to enhance mobile endpoint security control.
Security Intelligence: As cybersecurity attacks develop, advanced security solutions are important. Securing data and intellectual property against attackers is crucial for maintaining an organization's reputation. Security Intelligence's mission is to deliver actionable and comprehensive knowledge that decreases operational effort and risks. It makes advantages of cognitive AI power to automatically examine signs of compromise and gain vital insights. Artificial Intelligence and human intelligence are combined to produce cognitive security. It improves understanding of threat landscape, reduce false positives and so on.
INPUTS PROCESS OUTPUTS
Machine learning training
0 Dcnotejngand feature «etectfon
0 Machine learning model training
О
Constantly repeat^-?
nflfft с mdpom JTCmd
Diu COII1<<K1<
Fig. 1 The control scheme of mobile endpoint security CONCLUSION
Artificial Intelligence made a splash in cyberspace, promising to improve how humans interact with data. A substantial interaction between AI systems and human factors is required to increase the maturity of cybersecurity. The study provided a * comprehensive review of the most recent trends and challenges in cybersecurity as well as implementations of Artificial Intelligence on these challenges.
As the battle against cybercrime continues, AI solutions will become more successful than ever. Using automated techniques to fight hackers produces better outcomes than manual systems. Although it is hard to imagine a world without human resources fighting with cyber-attacks, recent developments in Artificial Intelligence will significantly decrease our reliance on their help.
Finally, AI can boost cybersecurity productivity and efficiency, but it requires appropriate skill sets and right integration.
References:
1. URL: Режим доступа: https://www.ibm.com/security/artificial-intelligence
2. URL: https://www.computer.org/publications/tech-news/trends/the-impact-of-ai-on-cybersecurity
3. URL: www.makeuseof.com/downsides-artificial-intelligence-cybersecurity/
4. Matthew N.,O. Sadiku, Omobayode I. Fagbohungbe, and Sarhan M. Musa, Artificial Intelligence in Cyber Security: Prairie View A&M University.
5. Artificial Intelligence in Cybersecurity, Nadine Wirkuttis and Hadas Klein.
6. URL: https://www.issquaredinc.com/insights/resources/blogs/pros-and-cons-of-artificial-intelligence-in-cybersecurity
7. URL: https://www.forbes.com/sites/louiscolumbus/2019/09/25/10-ways-ai-and-machine-learning-are-improving-endpoint-security/
