ARTIFICIAL INTELLIGENCE CRIME:AN OVERVIEW OF MALICIOUS USE AND ABUSE OF AI

INFORMATION TECHNOLOGY

This project provides a comprehensive overview of the growing concerns surrounding the malicious use and abuse of Artificial Intelligence (AI) in criminal activities. As AI technologies rapidly evolve and integrate into various sectors, they also become susceptible to exploitation for harmful purposes. Drawing on relevant literature, reports, and representative incidents, this study constructs a typology that delineates the multifaceted ways in which AI capabilities are manipulated for malicious ends. The primary objective is to shed light on the diverse activities associated with AI-related crime and the corresponding risks. The analysis begins by identifying vulnerabilities in AI models, offering insights into potential points of exploitation by malicious actors. Subsequently, the exploration extends to AI-enabled and AI-enhanced attacks, presenting a nuanced perspective on the risks without aiming for a conclusive and exhaustive classification. Specifically, the study proposes four types of malicious AI abuse, including integrity attacks, unintended AI outcomes, algorithmic trading, and membership inference attacks. Additionally, it highlights four categories of malicious AI use: social engineering, misinformation/fake news dissemination, hacking, and the deployment of autonomous weapon systems. This abstract concludes with a call for increased awareness, collaborative governance strategies, and policy development to minimize risks and mitigate the harmful consequences of AI-related crime. The interdisciplinary nature of addressing these challenges emphasizes the importance of collaboration among governments, industries, and civil society to enhance preparedness and resilience against the malicious use and abuse of AI.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.