Generative AI for Trustworthy and Compliant Cybersecurity in the Digital Landscape

Abstract

Generative Artificial Intelligence (GenAI) offers transformative potential for cybersecurity, enabling automation, adaptability, and intelligence against increasingly sophisticated threats. This paper examines the development, training, and deployment of GenAI models for threat detection, anomaly analysis, incident response, and compliance management in digital systems. We highlight the dual role of GenAI: as a tool to enhance security operations through self-healing, adaptive rule generation, and real-time monitoring, and as a mechanism to ensure regulatory compliance with evolving frameworks such as the AI Act, the NIS2 Directive, and the Cyber Resilience Act (CRA). Aligned with the principles of Trustworthy and Responsible AI, our approach integrates legal, ethical, and technical safeguards and leverages open-source tools and datasets to promote sovereignty and innovation. The proposed architecture emphasises interoperability, privacy by design, and documentation to ensure replicability and adoption across enterprises, providing a comprehensive view of how GenAI can strengthen cybersecurity ecosystems while maintaining compliance and transparency.