Cyber defence automation: Can AI outperform hackers?

Abstract

1. Growing cyber threats in the context of digital transformation require improved methods of automating cyber defence, in particular, through the use of artificial intelligence to detect and respond to attacks. The purpose of the study was to assess the potential of artificial intelligence in the field of cyber defence automation, and to determine its effectiveness in countering advanced cyber-attacks. The study was based on methods of comparative analysis, a systematic approach, and forecasting of cyber threats using artificial intelligence. For the first time, the effectiveness of using artificial intelligence in automated cyber defence systems was comprehensively considered in the context of the current challenges of the digital environment. An analytical review of the capabilities of artificial intelligence in detecting and predicting cyber threats using neural networks, security information and event management systems, and machine learning algorithms was presented. The potential and limitations of automated response systems were determined, and the risks associated with the possibility of using artificial intelligence by attackers themselves were outlined. Special attention was paid to the analysis of the role of the human factor in interaction with automated systems – it was shown that full automation does not guarantee security without the critical participation of specialists. Based on comparative and predictive analysis, practical approaches to the balanced implementation of artificial intelligence in cyber defence systems were proposed. It was proved that artificial intelligence can be a powerful tool for automating cyber defence, but to ensure a high level of security, it is necessary to consider potential risks and constantly improve systems. The practical significance of the study was to develop recommendations for implementing effective security technologies in various cybersecurity sectors