Abstract
The use of Artificial Intelligence (AI) in the banking sector can enhance the effectiveness and efficiency of banking services. However, it also carries the risk of misuse by irresponsible parties, potentially leading to significant financial losses. This research aims to identify and analyze the potential misuse of AI in banking systems and provide recommendations for risk mitigation.This study employs a normative juridical research method, incorporating case studies and legislative analysis. The findings reveal vulnerabilities in AI algorithm security within the banking sector and demonstrate that AI implementation can be exploited for fraudulent activities. These findings underscore the need for updates to procedural laws, particularly concerning the evidence system, to establish clear criminal liability for legal entities or individuals.The study concludes that regulatory adjustments related to the evidence system and the implementation of banking principles are essential to minimizing the risks of AI-based fraud.