Year: 2025
Type: article
Abstract: JISEM. Please publish. VenuReceived date:06-01-2025Revised:20-02-2025Accepted:09-05-2025Volume 2025Artificial intelligence (AI) has radically changed the cybersecurity environment in fintech, bringing sophisticated, adaptive threats that can pressure spilling disruptions. This paper examines how malicious AI agents can trigger disproportionately large effects through subtle, targeted disturbances—mirroring the "butterfly effect" of chaos theory. Employing a system dynamics approach, we simulate how AI-powered cyber-attacks use feedback loops, time delays, and interdependencies to spread across digital ecosystems. We reveal how AI capabilities – such as autonomous decision-making, adaptive targeting, and data weaponization – amplify attack vectors. Through recreation of various attack scenarios, we show the nonlinear characteristics of propagation and the significance of early detection and system resilience. The findings confirm that perimeter-based defense mechanisms are not sufficient, leading to resilience-driven approaches, cross-industry collaboration, and AI-enabled defense systems. This research offers theoretical insights and empirical models to increase preparedness against evolving AI-driven threats in more integrated systems.
Author: Uday B. Acharya
Institution: Hemchandracharya North Gujarat University
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4410322775