This article explores the potential of artificial intelligence (AI) in cybersecurity auditing and risk management within the context of ongoing digital transformation. Traditional approaches to information security auditing—based on manual data collection and periodic assessments—are increasingly insufficient for dynamic and large-scale digital ecosystems. They are limited in scalability, prone to human error, and lack the capacity for continuous monitoring. The integration of AI technologies allows for automated anomaly detection, proactive risk assessment, real-time decision support, and analysis of vast volumes of both structured and unstructured data, including event logs, network traffic, and audit reports. The study examines the application of machine learning and deep learning models in audit practices, including recurrent and convolutional neural networks, clustering algorithms, and natural language processing (NLP) techniques for detecting security policy violations. Particular attention is given to the concept of Network Situation Awareness, which enables the prediction of system behavior and potential threats based on historical and real-time behavioral data. In addition to technical achievements, the research addresses the ethical challenges associated with AI deployment in audits: algorithmic opacity, bias risks, privacy concerns, and difficulties in delegating decision-making to automated systems. The need for explainable AI (XAI) and the development of ethical guidelines for responsible AI use in cybersecurity audits is emphasized. AI is highlighted as a dual-use technology—capable of both defending against and facilitating cyberattacks. The article refers to real-world incidents, such as the use of generative models in social engineering and voice-based fraud. The aim of the study is to identify both the benefits and limitations of AI-powered cybersecurity auditing and to provide recommendations for the ethical and effective implementation of intelligent systems. The paper concludes that a hybrid model—combining AI automation with human expertise—is the most promising strategy for enhancing the accuracy, efficiency, and adaptability of cybersecurity risk assessment. This integrated approach is essential to improving cyber resilience in today’s volatile digital environment.
Obodiak et al. (Fri,) studied this question.