Ransomware attacks in the context of generative artificial intelligence—an experimental study

Abstract Ransomware attacks continue to be one of the biggest risks faced by both organizations and governments. In this experimental study, the author tested how criminals could use generative artificial intelligence (AI) for both the planning and implementation of ransomware attacks. It is found that criminals with very limited information technology (IT) knowledge may be enabled by chatbots using generative AI to carry out complex ransomware attacks. Furthermore, it is found that criminals with profound IT expertise but lacking other skills may be enabled by generative AI to draft more convincing phishing e‑mails. It is argued that the broad availability of generative AI could lead to an increase in both the number and the quality of ransomware attacks. Although previous studies have separately analyzed both ransomware attacks and generative AI, this article combines the two phenomena. The author uses criminological techniques and analyzes the potential use of AI from the perspective of a potential criminal. The risks identified in this article could serve as a foundation for further research in the fields of cybersecurity, IT law, and criminology.