OTENTIALS OF ARTIFICIAL INTELLIGENCE IN DETECTING AND PREVENTING PHISHING AND CYBER ATTACKS

Abstract

This paper examines the dual role of artificial intelligence (AI) in today’s cybersecurity landscape, highlighting its capacity to both enhance cyberattacks and support the development of effective defense mechanisms for information systems. The study analyzes the increasing complexity of threats driven by advancements in machine learning, natural language processing, and generative AI (GenAI), which enable attackers to automate, improve the accuracy of, and disguise attacks, including phishing, malware creation, and the use of deepfakes. This research explores how AI-based solutions can strengthen cybersecurity by detecting potential threats in real time through machine learning, NLP, and image recognition techniques. Special attention is given to the necessity of integrating AI with human oversight, emphasizing the importance of combining automated responses with expert analysis to effectively mitigate risks and adapt to emerging challenges. The paper reviews a range of modern tools and techniques used to execute AI-assisted cyberattacks, such as ChatGPT, WormGPT, FraudGPT, and Morris 2.0, demonstrating their capabilities in crafting convincing fraudulent scenarios and adaptive malicious software. In parallel, it examines AI-powered technological solutions designed to detect and counter cyber threats in real time. It describes the functionality of systems like Darktrace Antigena, Cylance Endpoint Security, Splunk, Exabeam, IBM QRadar, and Microsoft Sentinel, which leverage behavioral analysis, machine learning, and image recognition for proactive anomaly detection, automated incident response, and overall security enhancement. This study illustrates how AI is transforming cybersecurity by providing adaptive and proactive strategies to combat emerging cyber threats. By exploring contemporary AI applications, it demonstrates how AI can reshape cybersecurity by offering proactive and adaptive approaches to counter cyberattacks and protect sensitive data.

Ähnliche Arbeiten